New Subpanel

[David C. Roseman]

I'm no sparky either, but I know a bit about grounding and overcurrent protection. Art and Malcolm have quite correctly described what happens when a shorting current is either 1) carried back through the ground wire in the cable to the ground buss at the main service panel (which as everyone agrees, is bonded to neutral at that location), or 2) instead, carried directly to an earth ground far away from the main panel by way of a well-buried grounding rod. So long as the ground path is a good one, and the breaker properly sized to the conductor that suffers the short, the excessive amp draw will trip the breaker (or melt the conductor if improperly sized). It does not have to "loop back" to the ground buss at the main panel, thence to the earth ground established near that location by a dedicated ground rod or buried copper water pipe) in order to cause the breaker to trip, which is what I read Chris' comments to imply. Chris, I see now that is not what you meant, but I think Art and Malcolm took it that way as well. Sure, there must always be a loop for the current to flow. But it can get to earth ground by more than one path. Malcolm, thanks for adding "directly back to the main panel," but I think many of us probably knew that's what you meant. Chris, thanks for clarifying what you meant when you said "loop back."

This has been an interesting discussion.

[Wes Thom]

By the way, my shop and garage are part of the main house, all one structure. And by Julie's good diagram and David's comments it looks like I do not need a second ground rod on the sub panel.

Grounding serves many purposes. Posted were important facts mixed with a number of rumors - especially some based in experience without also understanding the many relevant functions.

First some accurate facts. Electricity is never same at both ends of a wire. Wires are electronic components. For example, both safety ground and neutral are bonded in the main box. But those same two wires are electrically different at each receptacle.

Safety ground, earth ground, chassis ground, digital ground, cold water pipe, analog ground, etc are all electrically different for the same reason and even if interconnectded.

Second, a facility must have only one earth ground. Failure to do this can create some human safety issues, create data communication issues especially among interconnected equipment, create ground loops, and can make surge damage easier.

Third, for human safety reasons, a sub-panel must have separate neutral and safety grounds. However if that sub-panel is part of an electrically separated building, then safety ground is replaced by an earthing electrode at the sub panel. Then any data connections between equipment in what are electrically considered two separate buildings requires special care. IOW interconnections from one 'side' must start at a main panel earth ground, then connect to the sub-panels earth ground before connecting to equipment powered by the sub-panel. This is made unnecessary if the entire building is one big single point earth ground.

Only some of the above is defined by code - which is only about human safety and is not about protecting equipment. Other functions of the earth and safety grounds (which are electrically different) involve reliable equipment operation including noise interference and transient damage.

For human safety, safety ground must connect current back to the main panel's bus bar to that panel's earth ground, and eventually to a neutral tap on the utilities transformer. Circuit breakers then trip to protect human life. That does not protect equipment.

Fourth, earth ground provides many functions. For example, it can avert safety problems if/when a neutral wire fails. It can avert a voltage difference between electrical safety grounds and (ie) concrete beneath feet (since concrete and other floor materials are electrical conductors). If installed to well exceed code requirements, then it provides THE most critically important protection from equipment destructive transients (ie a lightning strike far down the street to utility wires). Earth ground must have low resistance for human protection; and low impedance for equipment protection. Therefore earthing to protect equipment must both meet and exceed what is required by code for human safety.

This only summarizes many aspects of the different grounds. Essential to understand why grounding is done. This does not provide details and numbers necessary to better understand why we do what we do. It only summarizes what must be understood to appreciate the many reasons for both safety ground and earth ground.

If that sub-panel is earth grounded, then best is an earth ground conductor interconnecting both electrodes that is best buried in earth, is never inside or pass through any metallic material (ie bulkhead or conduit), has no sharp bends, no splices, and some other requirements all necessary to make that interconnection low impedance. The underlying principle here is equipotential - more commonly referred to as single point earth ground.

For human safety, that four wire interconnect (even 180 feet distant) is sufficient (as long as the installation does not compromise other code requirements). Meanwhile, safety grounds to water pipes is a human safety violation, is no longer permitted by code, and does not provide an effective earth ground. Pipes must be bonded and must not provide the earth ground.

Some basic principles apply including ground loops, low impedance, equipotential and conductivity, and the many underlying purposes of earth ground - that is electrically different from safety ground. Discussion should break down into two major and unique categories - human safety and equipment safety.

[Chris Padilla]

Nice, Wes. Thanks for your insight. On your point four (4), sometimes earth grounds (rod driven into the earth) on equipment can actually serve as a path for noise to enter equipment from events like lightning strikes (among other things) and disconnecting the earth ground fixes things. As long as the EGC (Equipment Grounding Conductor...the green/bare copper wire...ground wire) is connected, things should be fine, and the earth ground may not be needed.

I still don't agree that a sub-panel should need an earth ground. It must have an EGC or physics dictate that any 'safety current' on that EGC will never reach the mains to trip the breaker to de-energize something should shouldn't be. Are there instances in the NEC (or anywhere else) that say an earth ground for a sub-panel is fine in lieu of the EGC?

And just for completeness, I am not an electrician. I do however have an MSEE with a focus on electromagnetic theory. My job is a signal integrity engineer for high-speed circuits that run 10s of GHz. 60 Hz is a bit slower but the same rules of physics still apply.

[Wes Thom]

On your point four (4), sometimes earth grounds (rod driven into the earth) on equipment can actually serve as a path for noise to enter equipment from events like lightning strikes (among other things) and disconnecting the earth ground fixes things.

I don't know where I said that. However, earthing equipment can make that equipment a best (lightning) connection from cloud to earth; making appliance damage more likely. Transient protection is about connecting that current to earth on a path that is farther from the protected appliance. Single point earth grounding (not earthing separate equipment) puts everything at a common voltage.

Better protection means a lower impedance to earth - as short as possible and outside the building. And a higher impedance (while also maintaining a lower resistance) connection from the service entrance and its earth ground to equipment (ie increased separation between earth and equipment).

Best single point earth ground for a building is done for reasons similar to a motherboard ground plene. In both cases, equipotential is the objective. Then everything in a building or on a motherboard suffers from less stray currents, ground loops, and other resulting and problematic voltages.

60 Hz is why resistance is relevant. Transients that create noise or cause damage are much higher frequencies - why impedance is relevant.

[Art Mann]

Sorry Chris, but the breakers you are talking about tripping do not have a connection to either neutral or ground and have no way of detecting where the current is going. Fault current that is conducted through a short piece of wire to a good local double rod ground system will find less resistance than the path through 175 feet of wire and then going through a double rod ground system. You have this idea that the current must somehow make its way back to the main breaker and that just isn't true. Maybe you should stop thinking about electromagnetics for a little while and dig out your power transmission text. You are overthinking this.

[David L Morse]

Why is it that so many people think that dirt is a good conductor?

[Chris Padilla]

Sorry Chris, but the breakers you are talking about tripping do not have a connection to either neutral or ground and have no way of detecting where the current is going. Fault current that is conducted through a short piece of wire to a good local double rod ground system will find less resistance than the path through 175 feet of wire and then going through a double rod ground system. You have this idea that the current must somehow make its way back to the main breaker and that just isn't true. Maybe you should stop thinking about electromagnetics for a little while and dig out your power transmission text. You are overthinking this.

Art, this makes no sense and you don't have to keep apologizing. The breaker that would trip is the breaker serving the sub-panel. Did you watch the video I linked? It makes all this quite clear. As David questioned, the resistance of the earth is much much higher than a copper conductor. Please watch that video...it clearly explains the spreading resistance of the earth. The current flowing through the earth gets spread out among the earth...it is not focused like it would be in a wire. Current must return to its source...must flow in a loop. If you cannot understand or buy that...which is fundamental in electricity...I have no idea what to say. You must convince yourself that this is indeed true. I promise you that it is! LOL! I could post or point you to Maxwell's Equations or Ampere's Law and Kirchhoff's Loop or the Conservation of Energy but that is a lot of math...maybe searching on these you can find something that clicks for you. I bid you peace on the topic...I'm not sure what else to say.

[Mike Heidrick]

Ken, you run your four wires to the new panel's location yet from your main panel?

[Dan Hintz]

Why is it that so many people think that dirt is a good conductor?

If you pour a lot of water and salt in it, it's not too bad

[Wes Thom]

Why is it that so many people think that dirt is a good conductor?

Even concrete and linoleum tile can be a good conductors. But the word 'good' is subjective. Long before asking what is and is not a good conductor, one must first define parameters for that electricity.

Earth ground is a superb conductor for what we use it for. It was even used to carry telephone conversations. But like all conductors (including a copper ground plate inside computer motherboards), it has limits. Yes a voltage difference exists even end to end on that copper plate that is everywhere inside a motherboard.

Subjective reasoning is how myths and confusion get created. Perspective (numbers) are required in technical discussions.

[Chris Padilla]

As you say, Wes, the devil is in the details but if we are talking computers, that is a different class where wave theory should be considered as the wavelengths of the signals/electricity being considered there are roughly on the order of inches and less...comparable to the motherboard. 3 GHz has a wavelength of 10 cm (about 4") in air.

In discussing electricity within our homes, 60 Hz is a long wavelength (around 3,100 miles) compared to that vicinity and so circuit theory, more simplistic, is easier to use. The whole discussion I spawned is directly related to if 175' feet of earth/dirt/concrete/linoleum? being a better conductor than 175' of #8 or #10 or whatever copper wire. That lead to other discussions of how the EGC works (or doesn't work) and if current flows in loops or not or returns to its source or not.

[Wes Thom]

As you say, Wes, the devil is in the details but if we are talking computers, that is a different class where wave theory should be considered as the wavelengths of the signals/electricity being considered there are roughly on the order of inches and less...comparable to the motherboard.

Again, that lower resistance in the 175 foot safety ground is essential for human protection. Meanwhile earth ground is doing addition human protection. AND is essential for doing equipment protection.

The term is 'single point earth ground'. One reason why that is so important is that earth is so conductive - for that type of electricity - and is part of a solution that involves low impedance - not just low resistance.

His installation involves two major and different categories: human safety as defined by code and equipment protection that means exceeding code requirements even (and especially) for safety ground and earth ground.

So as to need not handle equipment as if in two separated buildings, the four wire connection to that sub-panel is recommended. Other grounding considerations apply. Those have not yet been discussed. A first post introduced how much there is to learn. Since wiring for human safety is easy. Wiring to protect equipment is harder and involves concepts little known to electricians such as E-M theory concepts. Two relevant terms are equipotential and conductivity. Even wire impedance can make or break equipment reliability especially in a facility where equipment is interconnected by signal/data wires.

[David L Morse]

Subjective reasoning is how myths and confusion get created. Perspective (numbers) are required in technical discussions.

Yes, let's look at some numbers.

Soil resistivity is typically in the range of 2,000 to 20,000 ohm-cm with some exceptions. Copper resistivity is 0.0000017 ohm-cm. So Copper is about a billion times more conductive than dirt.

[Wes Thom]

Soil resistivity is typically in the range of 2,000 to 20,000 ohm-cm with some exceptions. Copper resistivity is 0.0000017 ohm-cm. So Copper is about a billion times more conductive than dirt.

But copper is a 4 mm thick conductor. Earth is a many meters thick conductor. Don't ignore those numbers.

Soil is not being used for what copper is being used. Copper is being used for electricity where resistance is relevant. Earth is important for other currents with other parameters apply. You are mixing apples and oranges. To accomplish what earthing does is why earthing is sufficiently conductive. Then we may add other features (ie halo ground, Ufer ground, etc) to make earthing even more conductive.

So what is your point? Why are you arguing what was made irrelevant by previous posts?

[Tom M King]

This thread should be added to the stickies at the top of this forum. This is a fairly frequent question, and there is good education on it in this thread.