Remote building ground electrode - function?

[W@ttson]

To add to this discussion:

IEEE 142 (green book) Grounding of Industrial and Commercial Power Systems states the following reasons for system grounding.

1.3 Purposes of system grounding
System grounding is the intentional connection to ground of a phase or neutral conductor
for the purpose of:
a) Controlling the voltage with respect to earth, or ground, within predictable limits,
and
b) Providing for a flow of current that will allow detection of an unwanted
connection between system conductors and ground. Such detection may then
initiate operation of automatic devices to remove the source of voltage from these
conductors.
The NEC prescribes certain system grounding connections that must be made to be in
compliance with the code. The control of voltage to ground limits the voltage stress on the
insulation of conductors so that insulation performance can more readily be predicted. The
control of voltage also allows reduction of shock hazard to persons who might come in
contact with live conductors.

 

[W@ttson]

and elsewhere in IEEE 142:
4.1.3 Recommended acceptable values
The most elaborate grounding system may not perform satisfactorily unless the
connection of the system to earth is adequate for the particular installation. It follows,
therefore, that the earth connection is one of the most important parts of the whole
grounding system. It is also the most difficult part to design.
The connection to earth or the electrode system needs to have a sufficiently low resistance
to minimize the potential for shock to personnel who may be in the vicinity of equipment
frames, enclosures, conductors, or the electrodes themselves as a result of lightning
strokes or unintentional contact of conductors with higher voltage conductors and to limit
transient overvoltages.
The National Electrical Code® (NEC®) NFPA 701 for example, stipulates that the earth or
structural metal frame of a building shall not be used as the effective (sole) equipment
grounding conductor (EGC) and that a grounded system conductor shall be run to each
service.

 

[W@ttson]

After staring at this for a while now, and looking at a few IEEE publications I think the remote ground rods are about touch(or step potential) potential in the event of line to ground fault (not line to grounded metal).

IEEE 80 states the following:
4.1 Basic problem
In principle, a safe grounding design has the following two objectives:
― To provide means to carry electric currents into the earth under normal and fault
conditions without exceeding any operating and equipment limits or adversely affecting
continuity of service.
― To reduce the risk of a person in the vicinity of grounded facilities being
exposed to the danger of critical electric shock.
A practical approach to safe grounding thus concerns and strives for controlling the interaction of
two grounding systems, as follows:
― The intentional ground, consisting of ground electrodes buried at some depth below the
earth’ surface.
― The accidental ground, temporarily established by a person exposed to a potential
gradient in the vicinity of a grounded facility.

In the event that there is a rogue wire coming out of the subpanel in the remote building and it touches the ground, if there wasn't a ground rod there, the energy would have to flow into the ground through whatever distance of earth separate the main service and that remote building up the ground electrode, into the ground bus, into the Main bonding jumper, then into the neutral, back to the source. Given the resistance this may never trip a breaker. In this path back to the source, there might be potential difference in the earth creating a hazard for someone or something (cattle) in the area.

Now if the remote building had a ground electrode, and the same line to ground fault occurs, the energy this time would go into the ground then to the nearby ground electrode, then to the GEC at the remote building, which will go to the ground bus at the subpanel and subsequently into the EGC that was brought to the building. This will then in turn bring it back to that Main bonding jumper back at the main panel and off through the neutral. This is a much better scenario I would think.

I think Mike Holt does a terrific job showing some of this potential different in this video:

 

[kwired]

My thoughts:

#1. Just not real confident a few (more) rods will do anything meaningful to mitigate damage from a lightning event. I suppose I could be convinced if there were some facts or detailed scientific analysis on the subject.

#2. This should be deleted. This statement has been questioned by many members of this forum over the years.

#3. This is valid

#4. I think this should be deleted or reworded. It is true but I fear many read too much into it. You just create a fixed voltage reference by grounding a system. It gets rid of capacitive coupling that would give unpredictable and varying voltage readings to ground. The key is the "during normal operation" part - it all goes to hell during a fault. Also you don't need a low resistance to ground to get your stable ground reference. In the context of the specific situation of an outbuilding, the system is already grounded (ignoring ungrounded systems). If you have some difference of potential between dirt and the egc system at a remote structure, adding some more ground rods isn't really going to do anything, I mean maybe you could close it up some fraction of a volt, but that's about it.

I pretty much agree with all of that.

Earth is the reference, and it is generally the same potential everywhere. May have gradients near an electrode but otherwise is all at same potential.

In a separate building supplied by a feeder with a separate EGC, any voltage between the EGC and earth is a result of voltage drop on the service grounded conductor and not a difference in earth voltage at that location. Driving a ground rod here will not magically bring the EGC to earth potential, it will only do so for a very close distance to the rod as you will usually be lucky to get less than 10 ohms resistance out of the rod to earth connection.

Earth is a constant, making a connection to it is not so easy to do at a low resistance level.

CEE's, building steel, and water pipes usually have the most surface contact and therefore typically have lower resistance than a single rod or pipe electrode.

Even ungrounded services still must bond all non current carrying components and connect to a GES and run EGC's with the feeders and branch circuits. Those systems won't have the stray voltages a system with a grounded current carrying conductor may have, they can have stray voltages from capacitive effects though.

 

[kwired]

After staring at this for a while now, and looking at a few IEEE publications I think the remote ground rods are about touch(or step potential) potential in the event of line to ground fault (not line to grounded metal).

IEEE 80 states the following:
4.1 Basic problem
In principle, a safe grounding design has the following two objectives:
― To provide means to carry electric currents into the earth under normal and fault
conditions without exceeding any operating and equipment limits or adversely affecting
continuity of service.
― To reduce the risk of a person in the vicinity of grounded facilities being
exposed to the danger of critical electric shock.
A practical approach to safe grounding thus concerns and strives for controlling the interaction of
two grounding systems, as follows:
― The intentional ground, consisting of ground electrodes buried at some depth below the
earth’ surface.
― The accidental ground, temporarily established by a person exposed to a potential
gradient in the vicinity of a grounded facility.

In the event that there is a rogue wire coming out of the subpanel in the remote building and it touches the ground, if there wasn't a ground rod there, the energy would have to flow into the ground through whatever distance of earth separate the main service and that remote building up the ground electrode, into the ground bus, into the Main bonding jumper, then into the neutral, back to the source. Given the resistance this may never trip a breaker. In this path back to the source, there might be potential difference in the earth creating a hazard for someone or something (cattle) in the area.

Now if the remote building had a ground electrode, and the same line to ground fault occurs, the energy this time would go into the ground then to the nearby ground electrode, then to the GEC at the remote building, which will go to the ground bus at the subpanel and subsequently into the EGC that was brought to the building. This will then in turn bring it back to that Main bonding jumper back at the main panel and off through the neutral. This is a much better scenario I would think.

I think Mike Holt does a terrific job showing some of this potential different in this video:

What is the resistance of the grounding electrode though, as in resistance between the GEC and "earth"? If you get 5 ohms out of a rod (which is very good for a rod) and apply a fault voltage of 120 to it that is still only 24 amps of current that will flow. Will trip 15 and 20 amp OCPD's, but not on the instantaneous trip function, will leave you with voltage gradients and shock hazards for however long (probably minutes) it takes to trip on thermal trip function. Bonding and EGC's is what protects us from those events, the trip is nearly instantaneously if the current is high enough.

 

[ActionDave]

and elsewhere in IEEE 142:
4.1.3 Recommended acceptable values
The most elaborate grounding system may not perform satisfactorily unless the
connection of the system to earth is adequate for the particular installation. It follows,
therefore, that the earth connection is one of the most important parts of the whole
grounding system. It is also the most difficult part to design.
The connection to earth or the electrode system needs to have a sufficiently low resistance
to minimize the potential for shock to personnel who may be in the vicinity of equipment
frames, enclosures, conductors, or the electrodes themselves as a result of lightning
strokes or unintentional contact of conductors with higher voltage conductors and to limit
transient overvoltages.
The National Electrical Code® (NEC®) NFPA 701 for example, stipulates that the earth or
structural metal frame of a building shall not be used as the effective (sole) equipment
grounding conductor (EGC) and that a grounded system conductor shall be run to each
service.

If this is what engineers are learning about grounding/earthing it is no wonder they are so screwed up. It starts off saying that the connection to earth is the most important thing there is and then goes on to describe conditions that have nothing to do with earthing.

 

[kwired]

If this is what engineers are learning about grounding/earthing it is no wonder they are so screwed up. It starts off saying that the connection to earth is the most important thing there is and then goes on to describe conditions that have nothing to do with earthing.

And inspectors are brainwashed that grounding electrodes are about the most important thing in any installation as well. Haven't had one give a good reason why is so important though, just that it is code required and usually that if lightning is involved the worst case scenario is definitely going to happen.

Sorry but nearby lightning strikes - may help may not. Additional surge protection devices can help out, you probably will never know if they helped unless you noticed their indicator light stopped working after that specific event, then it might have offered some protection.

Direct hit, you have damages every time, some cases limited damages, some cases maybe the whole place burns down.

Build a faraday cage around your structure - best chances for little to no damages inside from a direct hit, but I wouldn't say it is impossible to have damages inside.

 

[W@ttson]

Here is the link to the Mike Holt Video that describes the ground rod issue. He goes on about lightning quite a bit in the beginning of where I linked the video but if you want to jump to 31:10 he describes the ground rod at the separate building.

Ends at 34:50

 

[W@ttson]

I pretty much agree with all of that.

Earth is the reference, and it is generally the same potential everywhere. May have gradients near an electrode but otherwise is all at same potential.

In a separate building supplied by a feeder with a separate EGC, any voltage between the EGC and earth is a result of voltage drop on the service grounded conductor and not a difference in earth voltage at that location. Driving a ground rod here will not magically bring the EGC to earth potential, it will only do so for a very close distance to the rod as you will usually be lucky to get less than 10 ohms resistance out of the rod to earth connection.

Earth is a constant, making a connection to it is not so easy to do at a low resistance level.

CEE's, building steel, and water pipes usually have the most surface contact and therefore typically have lower resistance than a single rod or pipe electrode.

Even ungrounded services still must bond all non current carrying components and connect to a GES and run EGC's with the feeders and branch circuits. Those systems won't have the stray voltages a system with a grounded current carrying conductor may have, they can have stray voltages from capacitive effects though.

After I posted that, I watched the Mike Holt video on stray voltage. Excellent video. It describes your above sentence to a T. Here is it:

End at 12:00
Toward the end of the clip above Mike admits of making the mistakes about using grounding to bring that neutral voltage to 0V in the circa 2002 Mike Holt material. He has since corrected this statement in his material.

All that grounding does to an electrode is energize the electrode.

Bringing an electrode to remote earth, driving it, connecting a new GEC for that remote earth/building, and then bonding an EGC to it from the main service will energize that remote earth GEC to the same potential as the main service. (In Mike's example in that video it is 3V). Essentially this is what is done for pools/cattle so that they don't feel the stray voltage differential. In that stray voltage video he goes on to explain that you are actually energizing that ground system as well as even the water in the pool.

He shows that pool water being energized to the same potential as the neutral earth voltage drop at his pool in this video:

(his neutral to ground voltage drop was 0.38V in that swimming pool video. )

 

[jaggedben]

Here is the link to the Mike Holt Video that describes the ground rod issue. He goes on about lightning quite a bit in the beginning of where I linked the video but if you want to jump to 31:10 he describes the ground rod at the separate building.

Ends at 34:50

You know it's interesting. He spends a good portion of the 3 minutes you're pointing to discussing that the NEC is not a lightning protection system. Then after that he spends some 20 minutes describing myths about grounding and all the ways in which it doesn't really help the way people think it does. Around the end of that section (before he starts talking about welding) he points out that the NEC doesn't require electrodes at lighting poles and there's no reason to put them there even though people do it all the time. Now, what is the difference between a lighting pole and a feeder fed structure? I mean, I'm not saying there's never a difference. The separate building could be a whole house with lots of metal piping systems and so on that might flash over with nearby lightning. But if it's say, just a detached garage 10ft from the house that has a subpanel with two circuits for garage door, lights and a couple plugs... I just don't know why in that case the NEC should require an electrode when it doesn't at a light pole. It seems to me that the NEC paints everything with a broad brush and there's some minority of cases where the electrode is justified, and the rest of the time it doesn't do any harm other than costing people money, and nobody can prove it's equally safe to not have it. So, it stays in the code.

BTW, my favorite part of the video starts at about 50:30.

 

[W@ttson]

You know it's interesting. He spends a good portion of the 3 minutes you're pointing to discussing that the NEC is not a lightning protection system. Then after that he spends some 20 minutes describing myths about grounding and all the ways in which it doesn't really help the way people think it does. Around the end of that section (before he starts talking about welding) he points out that the NEC doesn't require electrodes at lighting poles and there's no reason to put them there even though people do it all the time. Now, what is the difference between a lighting pole and a feeder fed structure? I mean, I'm not saying there's never a difference. The separate building could be a whole house with lots of metal piping systems and so on that might flash over with nearby lightning. But if it's say, just a detached garage 10ft from the house that has a subpanel with two circuits for garage door, lights and a couple plugs... I just don't know why in that case the NEC should require an electrode when it doesn't at a light pole. It seems to me that the NEC paints everything with a broad brush and there's some minority of cases where the electrode is justified, and the rest of the time it doesn't do any harm other than costing people money, and nobody can prove it's equally safe to not have it. So, it stays in the code.

BTW, my favorite part of the video starts at about 50:30.

I think the caveat about the ground rods is that it is just a ground rod with out it being connected to an EGC is the issue he is describing. In the samples he talks about the equipment becomes the path for the energy. If the additional ground rods are bonded to the overall Ground electrode system, then they are fine. It is as if you just added more ground rods in remote earth to lower the resistance of the GEC (making one really big ground electrode system). He shows that by having two ground electrodes in parallel driven at 50ft makes the overall resistance lower than the lowest one in that experiment video I posted (post #23).

Regarding the ground rod at the lightpole, what he is trying to highlight in the video (for reference here it is:

), is that there was an incident where lightpoles didn't have an EGC and caused a dog to die by some sort of contact (he says that it died when peeing on the lightpole). Then the officials erroneously thought that by taking a ground rod and driving it next to all the light poles and just bonding the light pole to the ground rod that it was safe now to touch in the case of a line to metal fault. This proved to be wrong when the second dog died in 2000. Then Mike got involved, told them that the ground rods don't do anything since its not enough current being drawn to trip a breaker which is supplying the power to the light pole. He then made them run an EGC to all the poles, but they still drive the ground rods because of the "we've been doing this forever" reason.

I would say that if the ground rod is bonded to the EGC, then you are just creating a really large Ground electrode system and there is no harm no foul. The ground resistance at the service will be really low. In the NEC 250.121 does say that an EGC shall not be used as a GEC but there is an exception (basically, if you treat the EGC like a GEC and follow all the requirements for a GEC then you can use a EGC for a GEC and more importantly serve the function of both).

 

[W@ttson]

So I want to summarize everything here since there was a lot of discussion. After taking in everyone's very helpful input, looking at some IEEE standards, and a few Mike Holt videos I get the following for where we started this post:

What is the function of the ground electrode at the detached remote building?

The NEC has the following listed:

1. limit the voltage imposed by lightning
2. limit line surges
3. limit the voltage imposed by unintentional contact with higher-voltage lines
4. stabilize the voltage to earth during normal operation

From the Mike Holt video he seems to focus mainly on discussing lightning, though he does show a slide about sources that can produce similar transients/surges that are induced into the electrical system of a building and need to be dissipated back to ground. The book is from the 1940s and it lists the following as additional sources:
1. Static
2. Physical contact with a higher voltage system
3. Resonance effects in series inductive-capacitive circuits
4. Repetitive intermittent short circuits
5. Switching surges
6. Forced-current zero-current interruption
7. Autotransformer connections
8. Lightning

In the Mike holt video he talks about how if lightning strikes nearby it will induce a voltage in the wiring of the building and that energy needs to dissipate back to the ground through the electrode. This could be a lightning strike, or a line to ground fault of the utility, or any other of the 8 above.

There is debate whether it really is as effective and practical to drive ground rods to help in the situations above. For example, if lightning strikes nearby and induces a voltage on the electrical system in the building, is a driven ground rod going to dissipate enough energy to really save anything?

I don't know, but the above is the reasoning behind the ground electrode to earth and the reason why the remote building needs its own is discussed in the Mike Holt video. It is because the requirement for the GEC is for it to be as short, direct, and free of loops as practicable since in the event of a high frequency induced transient, increased length in the GEC increases the impedance drastically, making it less effective. That is why the EGC back to the main service ground electrode is not sufficient. Maybe it might be if the detached building is 5 ft away but may not be if the building is 100ft away, the NEC draws the line at "detached building". If it is detached then you got to follow the rules for ground electrodes no matter if it is 5 ft or 100 ft away.

Thank you to everyone for there insightful input, if any of the above summary still missed the mark, please correct me anywhere that you see fit.

 

[W@ttson]

One thing to add to the above is, why have grounding in the first place (i.e. why have that electrode at the source in the first place). You can have an ungrounded system and still have bonding to clear faults.

Well, according to the Mike Holt video, they found out that there were a lot of issues with ungrounded systems when you had the over voltage / transient conditions (flashovers). A lot of those were mitigated when adding a ground to the system and provided "system grounding" instead of just "equipment grounding".

Here is the video from Mike Holt:

In this part of the video he shows how the induced voltage supposedly gets dissipated to ground through the ground rod.

 

[kwired]

With MGN distribution, which is what a lot of POCO's use at the distribution level on their systems, you have thousands of rods and other electrodes on the system and can come up with grounded conductor that is relatively close to earth potential. Problem is they do use said conductor to carry current on the system, and that still leads to stray voltages due to voltage drop because the conductors are carrying current during normal operation.

This part of why equipotential bonding is critical in art 547 and 680 applications. The equipotential bonding isn't bringing the items being bonded to earth potential, it is bringing the items bonded to same potential as the grounded conductor so that animals/people can't touch an item that is at different potential than what they are standing on or immersed in.

 

[W@ttson]

With MGN distribution, which is what a lot of POCO's use at the distribution level on their systems, you have thousands of rods and other electrodes on the system and can come up with grounded conductor that is relatively close to earth potential. Problem is they do use said conductor to carry current on the system, and that still leads to stray voltages due to voltage drop because the conductors are carrying current during normal operation.

This part of why equipotential bonding is critical in art 547 and 680 applications. The equipotential bonding isn't bringing the items being bonded to earth potential, it is bringing the items bonded to same potential as the grounded conductor so that animals/people can't touch an item that is at different potential than what they are standing on or immersed in.

Exactly right. Neutral voltage drop causing a neutral to earth potential difference at the premises. Makes the cows not want to drink water or produce milk. Mike holt calls it "energizing" the water/rebar cage around the pool/livestock keeping areas to the neutral to earth voltage.

 

[kwired]

Exactly right. Neutral voltage drop causing a neutral to earth potential difference at the premises. Makes the cows not want to drink water or produce milk. Mike holt calls it "energizing" the water/rebar cage around the pool/livestock keeping areas to the neutral to earth voltage.

So many confuse this with "grounding". it is not grounding, it is bonding and does exactly what he said there. The key trick is not to let any "holes" in the grid, or you will have shock potential near those holes. It works for same reasons birds on bare overhead lines don't get shocked, you have to contact something of a different potential to be shocked.

I'm not necessarily explaining this just to you, I think you get it. Just posting it for anyone reading that still don't get it. Questions about this seem to come up pretty often.