Why a grounding electrode at sub-fed structures?

[Curious nerd]

Smart $ raises a point:
"Don't forget that the grounded conductor is bonded to the GES. That in itself helps 'stabilize' the system voltage, which is the point of bonding the GES. The downside is, yes, there can be voltage gradients across the system... but ask yourself, what is the alternative? Do you think fewer electrodes will change the voltage gradients imposed on conductors from a nearby lightning strike?"

There's a lightning expert with a PhD on this site, and if he speaks up I will fall silent.

In the meantime, if I've understood what I've read, multiple grounding locations means opening the door to ground loops. "Fewer electrodes" makes things worse, but designing with fewer points of attachment to the GES reduces the chance of damage. Ideally there should be only one point.

To put in another way, if you don't feed step potential into the conductors by grounding them at distant points, then you don't have the "voltage gradients imposed on conductors from a nearby lightning strike".

"What is the alternative?" Smart $ is asking a great question here. There's a terrible downside to a system-wide shared grounding point in the main structure, the kind of system I've seen advocated. Imagine lightning striking the detached building. Running that current inside the main building on its way to ground invites disaster.

 

[jaggedben]

Definitely a valid concern - then consider potential for arcing and flashover, especially if EGC is routed through/over/around building to connect to "auxiliary" GR for other equipment (telephone service, standby generator).

This is precisely why Mike Holt called for a temporary interim amendment to remove 690.47(D) from the 2014 code. The code required the auxiliary electrode for PV arrays but by calling auxiliary it only required an EGC connecting it to the main GES. And because we're dealing with solar, the likelihood of the EGC passing over the building is very high.

No TIA was passed, AFAIK, but it looks like the 2017 code will permit array electrodes, not require them.

 

[AJElectric]

In the meantime, if I've understood what I've read, multiple grounding locations means opening the door to ground loops.

That's my theory, but don't take it from me. I'm trying to figure out why ground rods all over the yard is a good idea. I too would welcome the input of a lightning expert.

There's a terrible downside to a system-wide shared grounding point in the main structure, the kind of system I've seen advocated. Imagine lightning striking the detached building. Running that current inside the main building on its way to ground invites disaster.

That's the thing - I don't see what the downside is because I'm thinking that if lightning strikes the remote structure itself, a ground rod or even 1000 ground rods tied to the EGC will offer no measurable protection.

To summarize, in my theory adding the GE at a remote structure is a means of picking up the voltage gradient from a nearby strike and giving it a shortcut to every GE at every structure - not cool. Meanwhile in the event of lightning strike upon a structure the GE doesn't offer any real advantage, a significant amount of energy will be carried across the EGC's all over the system anyway.

 

[AJElectric]

This is precisely why Mike Holt called for a temporary interim amendment to remove 690.47(D) from the 2014 code. The code required the auxiliary electrode for PV arrays but by calling auxiliary it only required an EGC connecting it to the main GES. And because we're dealing with solar, the likelihood of the EGC passing over the building is very high.

No TIA was passed, AFAIK, but it looks like the 2017 code will permit array electrodes, not require them.

I get the impression from his many training videos that MH is generally anti-GES-all-over-the-yard. He's definitely opposed to more than one at a single house, and 690.47(D) is where he finally drew the line and called out the BS because yes, that is especially dangerous considering a 14AWG wire on the roof could connect the two GE's together.

 

[Curious nerd]

I'll try again, maybe this will be clearer.

Imagine that there's no ground rod/plate/pipe/Ufer/etc. at the detached structure. Lightning hits an antenna or something on the detached structure. Lightning connects with the EGC. Of all the paths it will take to ground, the best one for it is now shed->EGC->main structure panel->main structure ground rods/plates/pipe/Ufer/etc->dirt.

Which means that the single-point-of-grounding design idea can lead to problems if implemented without a lot of thought.

 

[Tony S]

I recently designed an earthing (grounding) system for stables. The contractor was complaining of the cost and work involved with multiple rods until I asked how much a race horse was worth.

 

[Smart $]

I'll try again, maybe this will be clearer.

Imagine that there's no ground rod/plate/pipe/Ufer/etc. at the detached structure. Lightning hits an antenna or something on the detached structure. Lightning connects with the EGC. Of all the paths it will take to ground, the best one for it is now shed->EGC->main structure panel->main structure ground rods/plates/pipe/Ufer/etc->dirt.

Which means that the single-point-of-grounding design idea can lead to problems if implemented without a lot of thought.

It's not the grounding system that causes a problem... whether the new way or the old way (no EGC, N-G bond at each structure), or the number of electrodes... directly.

What causes a problem is when there are potential indirect paths in which current may pass through a person, animals, and/or equipment.

 

[AJElectric]

... Of all the paths it will take to ground, the best one for it is now shed->EGC->main structure panel->main structure ground rods/plates/pipe/Ufer/etc->dirt...

Understood, but how much of the energy can GE at the remote structure carry to ground? - what about the other thousands of kW energy that went "everywhere" anyway?

I recently designed an earthing (grounding) system for stables. The contractor was complaining of the cost and work involved with multiple rods until I asked how much a race horse was worth.

Do you mean an equipotential ground plane similar to a swimming pool, or is this a lightning protection system?

It's not the grounding system that causes a problem... whether the new way or the old way (no EGC, N-G bond at each structure), or the number of electrodes... directly.

I concur - new or old way - a conductor to earth is a conductor to earth.

What causes a problem is when there are potential indirect paths in which current may pass through a person, animals, and/or equipment.

I don't quite follow. Do you mean with respect to touch potential in the event of a fault/stray voltage, or the dangers of a lightning strike?

 

[jaggedben]

Just my opinion...

There are dangers that can derive from the electrical supply when conductors and/or equipment that are supposed to be grounded are not adequately grounded.

Then there are dangers related to lightning energy finding easy paths along electrical conductors and installations.

The two types of problems are totally different categories, and the solutions are more or less diametrically opposed. To a large degree the former can be used to justify more and more grounding everywhere, until you start creating problems with the latter. Those who live in lightning prone areas should be more concerned about excessive and improper grounding.

 

[Smart $]

...I don't quite follow. Do you mean with respect to touch potential in the event of a fault/stray voltage, or the dangers of a lightning strike?

Both, and any other.

 

[jwnagy]

Grounding electrode for detached building with subpanel

Grounding electrode for detached building with subpanel

Does anyone remember in which code cycle the requirement for a GE for a detached building with a subpanel was introduced?

 

[jumper]

Does anyone remember in which code cycle the requirement for a GE for a detached building with a subpanel was introduced?

It has been a code rule for a very long time. What is new is the requirement for a 4W feed with an EGC, 2008.