Electrical Theory Question: Grounding Rod On Detached Structures

[inq]

I'm trying to figure out exactly why the NEC requires separate grounding rods for detached structures (250.32). Mike Holt has made it clear in various comments and videos that its bad practice to install grounding rods unless its absolutely necessary, or required by code. While grounding rods can be an effective safety measure, they also can cause harm to people and equipment. If a large voltage source (such as a lighting strike) hits near them, they can provide a path for this current to get to equipment and destroy it or harm anyone touching it. They can effectively expand the area of danger for any electrical system.

I've thought about this and looked online, and the only explanation I am finding is that if a lighting strike has to travel back through a long route to the main panel where the ground wire is bonded to neutral bus it could generate so much heat that it could start a fire. Therefore, for "detached structures" we install grounding rods to allow that current to bypass the return path to the main panel and go directly to ground.

What is confusing me though is why the code doesn't mention this. No where does it stipulate how far away the detached structure has to be or how long the return path to the main panel is.

Is this really the reason we install grounding rods for detached structures? Am I missing anything?

 

[petersonra]

The reason is that somehow it got in the code and the current panels are afraid to take it out. Just like they are afraid to take out the pound two rods or pay big bucks to prove less than 25 ohms rule.

 

[electrofelon]

The reason is that somehow it got in the code and the current panels are afraid to take it out. Just like they are afraid to take out the pound two rods or pay big bucks to prove less than 25 ohms rule.

I agree. There is lots of stuff in the code that no one seems to know the justification of why it was put in, and the code panels are strangely extremely reluctant to remove or revisit these requirements.

 

[don_resqcapt19]

One of the issues with removing things that have been in the code for some time, is the requirement for a very solid technical substantiation to support the removal...for the issues in this thread, that would probably require actual testing.

 

[Dennis Alwon]

It is also better, imo, to have a rod at the structure so that any lightning, surges, etc go thru the rod rather than back thru to the house panel and then to the grounding electrode conductor's.

 

[electrofelon]

One of the issues with removing things that have been in the code for some time, is the requirement for a very solid technical substantiation to support the removal...for the issues in this thread, that would probably require actual testing.

But if no one can think of a reason for the requirement, and there probably never was one to begin with, what sense does it make it require a substantiation for removal? Thats what drives me crazy about the code writers is they often bring things into the code with no substantiation or evidence (like many of the expanded GFCI requirements or the surge protector) but try to get something changed or removed and its frequently "you havnt provided any evidence"

 

[tom baker]

But if no one can think of a reason for the requirement, and there probably never was one to begin with, what sense does it make it require a substantiation for removal? Thats what drives me crazy about the code writers is they often bring things into the code with no substantiation or evidence (like many of the expanded GFCI requirements or the surge protector) but try to get something changed or removed and its frequently "you havnt provided any evidence"

The testing and reasons for ground rods, etc was done perhaps 100 years ago, it would take some research to find the reasons. One reason has to do with grounding of electrical systems, originally ungrounded, there were fires from overvoltage, grounding was done. This reason is well known and documented. Soares book on grounding would be a good reference for you to start on your PI documentation

 

[don_resqcapt19]

But if no one can think of a reason for the requirement, and there probably never was one to begin with, what sense does it make it require a substantiation for removal? Thats what drives me crazy about the code writers is they often bring things into the code with no substantiation or evidence (like many of the expanded GFCI requirements or the surge protector) but try to get something changed or removed and its frequently "you havnt provided any evidence"

That is just how it works with the current rules for changing the code.
As far as the GFCIs, most have had some type of body count as a substantiation, whether that was sufficient or not, is very subjective.

 

[inq]

It is also better, imo, to have a rod at the structure so that any lightning, surges, etc go thru the rod rather than back thru to the house panel and then to the grounding electrode conductor's.

I'm still confused as to what scenario(s) this is guarding against. Since the grounding bus and the neutral bus are required to be completely isolated in a feeder panel. Wouldn't the lighting just travel back to the main panel anyway? It seams as though this is only guarding against a direct hit to the panel itself — and a grounding rod would make that risk even greater.

 

[inq]

The testing and reasons for ground rods, etc was done perhaps 100 years ago, it would take some research to find the reasons. One reason has to do with grounding of electrical systems, originally ungrounded, there were fires from overvoltage, grounding was done. This reason is well known and documented. Soares book on grounding would be a good reference for you to start on your PI documentation

There is an excellent write up of the history of residential grounding requirements here: https://www.epri.com/research/products/1005490

 

[Sea Nile]

It's not that multiple ground rods are bad,. It's just that you don't want to bond them to your house at more than one point.

Because lightning creates a voltage gradient you will have a massive difference of potential between the ground rods. And if you have your neutrals bonded to different rods at different locations, the voltage difference of the gradient will happily take the short cut of running through your conductors.

The reason a detached structure is ok to be have it's own ground rod is that the short cut the lightning takes will be outside your house.

At least that's my understanding.

It's the exact same reason you have step potential hazards around downed power lines. If you take a big step, the voltage gradient will travel up one leg and down the other. So you have to bunny hop or take baby steps. You don't want to have great distances between points when it comes to voltage gradients. But if the two points are both outside, little chance of starting a fire or electrocuting someone.

 

[Sea Nile]

Also, a house is naturally grounded in many points from gas lines, concrete pads, water lines etc. It's impossible to isolate your ground. But you can isolate your neutral to one point. So multiple ground rods around your home are meaningless. The only thing that matters is your neutral in bonded to the ground at one point and only one point

 

[kwired]

That is just how it works with the current rules for changing the code.
As far as the GFCIs, most have had some type of body count as a substantiation, whether that was sufficient or not, is very subjective.

The body counts were back in the 70's and 80's when GFCI's were still relatively new. Most changes since about the 96 code that had a body count attached to them were based on a count of just one, and possibly had other issues that contributed, like improper EGC conditions that likely would have prevented the body count if had been done correctly.

And of course manufacturers never use such examples to push a change so they can sell more their product either, you can bet most the changes have been initiated by a manufacturer or their representatives in this particular area.