New Subpanel

[Art Mann]

Maybe I missed it, but I don't think Ken's second question has been answered: Whether he needs a second ground rod for the subpanel, in addition to the question about bonding ground to neutral at the subpanel. Art, Julie and the other bona fide sparkys/EEs can answer this better than I, but it's my understanding that this depends on whether his shop is in the same building as the main, or in a separate structure. Here locally, a subpanel in a separate building requires it's own ground rod, while it does not if in the same building.

My advice to establish a second earth ground was based on the assumption that the sub panel is far away. The sub panel in my shop is located 175 feet away from the main panel. If the sub panel is close, it makes a lot more sense to just use the ground from the main panel.

[Julie Moriarty]

Grounding is probably one of the most controversial issues in electrical construction. When I was setting up services for cell sites in the 90's, I had to go to just about every municipality in the Chicagoland area and meet with the inspectors to explain what we were doing. Cell sites have their own grounding requirements and most inspectors were clueless about this new technology. When they started failing our work, I decided to be proactive and meet with the inspectors before the work began.

Grounding was a major part of our discussions. Some inspectors wanted the service to be grounded to a cold water pipe and a ground rod. Others said one or the other, never both. We have theories about what goes on in electrical grounding but I don't think anyone really knows 100% for certain. I've seen electricity do some pretty weird things no one could have predicted.

In the end, the easy answer is to follow your local code. If in doubt, call the inspector. If anything goes wrong, blame them.

[Art Mann]

Some consumer electronics and circuit protection devices specify that the resistance from the ground bar to earth ground be less than 25 ohms or something lIke that. I don't know exactly how that is measured.

[Dan Hintz]

We have theories about what goes on in electrical grounding but I don't think anyone really knows 100% for certain. I've seen electricity do some pretty weird things no one could have predicted.

Not sure I agree with that first part, at least as it's written. Where people get into trouble is when they try to ascribe perfect grounding properties to every type of ground. You need to know the reason behind grounding the item in question so you can make an informed choice.

For example, audio devices are often grounded at both ends... right up until ground hum starts, and then the immediate response is to open one end. It works (sometimes), but that means you've put a band-aid on without truly understanding the root cause. Grounding for lightning protection (okay, fine, lightning damage limiting ) is a totally different animal and doesn't follow the same rules. And again when it comes to grounding building electrical for safety.

Wires aren't perfect conductors, Faraday cages aren't perfect EMI isolators, coax/Ethernet shielding isn't perfectly noise-immune, etc. but it makes people feel better to think of it that way, so they make improper choices.

[Chris Padilla]

My advice to establish a second earth ground was based on the assumption that the sub panel is far away. The sub panel in my shop is located 175 feet away from the main panel. If the sub panel is close, it makes a lot more sense to just use the ground from the main panel.

175 feet traveling through the earth to complete its loop or 175 feet traveling through a #8 or #6 (or whatever wire) to complete its loop: which path has the least resistance and which path will actually trip the breaker as the ground is designed to do in the event that a hot leads touches metal it isn't supposed to?

The answer is the wire...not the earth. An earth ground in this case has WAY too much resistance to ever reach the main panel.

[Chris Padilla]

Wires aren't perfect conductors, Faraday cages aren't perfect EMI isolators, coax/Ethernet shielding isn't perfectly noise-immune, etc. but it makes people feel better to think of it that way, so they make improper choices.

Neither is the Earth a perfect conductor.

[Art Mann]

175 feet traveling through the earth to complete its loop or 175 feet traveling through a #8 or #6 (or whatever wire) to complete its loop: which path has the least resistance and which path will actually trip the breaker as the ground is designed to do in the event that a hot leads touches metal it isn't supposed to?

The answer is the wire...not the earth. An earth ground in this case has WAY too much resistance to ever reach the main panel.

There is no current loop in the ground line! All current travels from line to line or line to neutral. You need to have a ground sufficient to fault conduct current to earth ground. It dosen't matter where you develop it. In case you didn't realize it, the ground terminal at the main panel doesn't conduct any current either. If you ever have a high resistance or open circuit in the neutral line going back to your transformer from the main panel, you will see what I mean.

The reason to develop a local ground at a remote location is simple. Copper is expensive.

[Chris Padilla]

What happens when a hot lead connects to the equipment ground? It has current in that situation, right?

[Robert Payne]

As Art Mann pointed out, you must have a neutral to carry the current back to the source. Hot to Ground will show a voltage potential, but as soon as a load is placed between them, the load will not function. Certainly there is current that will pass to Ground, but it is shorting current. I recently had a situation with a broken Neutral at a neighborhood sign with lights that would not function; the voltage was measurable, but the lamps would not come on. Turned out to be a bad neutral feeding the meter by the power utility.

[Chris Padilla]

The safety ground is also there to carry current back to the main panel. The neutral and the ground are bonded together at the mains precisely for that reason. This isn't really about a functioning load but more of a safety situation. If the metal casing of a motor is contacted by a hot lead, it becomes energized and therefore a hazard if a human touches it and it is not grounded. If it is grounded via the equipment ground wire, the current will head back to the main panel and trip the circuit breaker thus removing the hazard. If you, instead, drive a ground rod into the ground and tie the sub-panel to that, the current in this hazardous situation will never get back to the mains to trip the breaker. The path through the ground rod back to the ground rod at the panel is simply too resistive and that breaker will never trip.

[Art Mann]

What happens when a hot lead connects to the equipment ground? It has current in that situation, right?

'

Yes, but it does not have to conduct current back to the center tap of the transformer (neutral). There is no loop. The safest ground is one that has the lowest resistance path to earth ground. I have experience with grounding problems dating back to when I owned a house on a lake and a floating aluminum boat dock. I installed a 20A circuit from the house to the dock to operate a boat lift. I electrically connected the dock to to the ground wire that came from the house. The first time I jumped in the water after installing this outlet, I felt a tingling sensation as I approached the dock so I swam to shore instead. This phenomenon didn't go away until I installed an earth ground at the shore next to the dock ramp.

You are not correct in stating that the ground is there to carry current back to the main panel. The ground at a remote sub panel doesn't carry any current at all unless it is wired wrong or you have a short. A person is protected by earth ground because it provides a much lower resistance path to ground (~25 ohms) than a human body and therefore, no current flows through the person. It is current, not voltage, that kills people. You can easily get killed by much less current than it takes to trip a 20A breaker (ground fault interrupter excluded). People have been killed through house wiring where the breakers were correctly installed and even a 15A breaker never tripped.

[Malcolm McLeod]

I'm not a sparky, but have to go with Art here (Sorry, Chris!)

My practical experience on such is that a ground exists to protect people. In simplistic terms, a machine or motor's ground path just has to have a lower resistance than the person who touches it. (Kind of like 2 guys running from a bear, 1 guy doesn't have to outrun the bear, only the other guy.)

If a motor frame is 'hot' due to a short and it is not grounded, whether running or not, the 'flow' of current may be normal, and so not cause the breaker to trip. But it IS still dangerous; a person (who may be grounded) walks up and touches it, person completes path to ground, person receives current flow, person gets zapped. Bad news. (Depending on Mr. Zap's bodily resistance to ground, the CB may still not trip.)

If the motor frame is properly grounded (low resistance) to anything, then the short will create current flow to ground. A person who touches it is now just a parallel path to ground (hopefully a high resistance one!). In a perfect world, the CB would be a GFCI, detect the abnormal flow to ground, and trip off - long before anyone touched it.

But the ground does not have to loop back to the main panel. Any ground will work. Locating it in the main panel is just convenient, cheap (hopefully), and standardized so it is easy to find (for future sparky work and inspectors).

[Chris Padilla]

Current ALWAYS makes a loop. It ALWAYS goes back to its source. This is a fundamental law of circuits and electromagnetic theory...Kirchhoff to be exact. So if we have a hot lead on the ground, then the current generated there will try to go back to the source...and in the case of a sub-panel this will always be the mains. My point, Gentlemen, is that the path to complete the circuit is much lower resistance through a copper wire than through the earth. And the ground wire is there for safety to complete that circuit to trip the breaker to deaden the motor's case, for example. The ground wire and the neutral are one and the same at the mains. Current will flow to it there and trip the breaker due to the short circuit. It has to. Current must complete the loop. Ground is not some magical place where current suddenly goes and dissipates and is gone forever. Electromagnetic theory doesn't work that way. Current always flows in a loop...this is fundamental. https://www.youtube.com/watch?v=qNZC782SzAQ And when you're done viewing this, check out a few more of Mike's videos. I think you'll find them educational.

[Steve Meliza]

I'd be surprised if your local code requires you to add a new ground rod for the sub panel considering it sounds like it will be located a few feet or less from the existing main. But as an example of the curves that inspectors can throw you, mine wanted the ground bar in the sub panel bonded to the enclosure (the metal box of the sub panel) and did not consider the mounting screws for the grounding bar to be sufficient. I had to grind paint off of the box, drill and tap a hole, and install a clamp with a ground strap running from it to the ground bar. My father, a licensed electrician in the area nearly his entire working life, had never heard of an inspector requiring that before.

On the other hand, the licensed electrician that was hired by the previous home owners to install a new main panel failed an inspection because the panel was below the surface of the wood paneling of the wall instead of being flush or proud of it. Even I know better than that. So to those who just to say "Hire an electrician" as their only advice I say "Fooey!"

[Malcolm McLeod]

My bad. You are correct about loops and I should have typed "loop directly back to the main panel".