Gate/Fence Grounding

[skeshesh]

Good day everyone.

I can't find a good reference for this. The NEC does not seem to have anything specific on the subject. I'm looking over a job a co-worker did before retiring and he's specifying 1"PVC with #2AWG being dragged from post to post and connected to a ground Rod. I'm not sure why he did not just use a bare conductor in that case to establish a ground ring and avoid using a ground rod, but that's not the issue.

The details show a typical Burndy post ground clamp connector with a 2/0AWG. I've often seen 1/0 or 2/0 specified for connecting the posts but have not done any calculations. The fence is not electrified and has no electrical provisions (access control, etc.). I looked up the calculations for determining the KCMIL for the ground but the simplified version requires short circuit levels and time of fault which I don't have and the longer formula requires additional information (resistivities at different temps, etc).

The guys installing this are questioning the size of the grounding conductor and I want to have a solid understanding of what's going on before I give them an answer. Any comments or referenced would be much appreciated.

 

[qcroanoke]

Is this fence around a communications site? (cell phones, police radios and such)
They are in their own world and a rule of thumb is if it doesn't move ground it!

Motorola has a standard that covers the whole building that is called R56. It is quite through in its requirements.

 

[skeshesh]

It is not. In addition, it is not substation fencing or any special application that may require other considerations. As stated it was not my project so I'm not very familiar with it but I believe it's just a regular chain link fence around some HVAC equipment and pumps (those were all existing).

 

[John120/240]

Is the fence likely to become energized ? Would not the metal fence post set

in the earth or concrete be already grounded ? Granted additional grounding

can't hurt but is it necessary ?

 

[skeshesh]

It is not easy for me to answer that with confidence, but I'd say it's not very likely to become energized. There are pumps and conduits around, but it's not like it's a fence sitting underneath a power line or one around a substation. Your comment about the fence posts has theoretical merit, but I'd say that placing a rod or ground ring would be necessary if you are planning to ground the fence. In any case, the client's design standards require it.

 

[qcroanoke]

It is not easy for me to answer that with confidence, but I'd say it's not very likely to become energized. There are pumps and conduits around, but it's not like it's a fence sitting underneath a power line or one around a substation. Your comment about the fence posts has theoretical merit, but I'd say that placing a rod or ground ring would be necessary if you are planning to ground the fence. In any case, the client's design standards require it.

If the client's design standards require it, I say just do it.Wether it makes sense or not......

 

[skeshesh]

If the client's design standards require it, I say just do it.Wether it makes sense or not......

My question wasn't really about whether it makes sense or not. I think my original question was not stated very clearly as I wrote it in a hurry. Another attempt: the original designer shows on the site plan a ground rod and 1"PVC-#2GND running around the fence and being connected to a fence post every some-odd feet. I'm trying to figure out how the conductor size for this application is determined.

 

[WIMaster]

I would believe they want to have a lightning protection system?
NFPA 780 essentially says that the smallest stranded conductor used on structures under 75' high should be a #2. (I am over simplifying).
I do not understand the 1" PVC, but I can understand the ground ring.
The post locations are usually set in concrete providing a small ufer ground at each post and bonding the fence helps to provide a faraday cage.
You will want to bond all of your available grounding electrodes together with a minimum of a #2.

 

[skeshesh]

WIMaster, thanks for the response, but this is not a lightning protection system. It is intended to ground the fence (metallic). This is NOT a ground ring using #2 as the ground rod is the grounding electrode, and like you stated the ground ring would not be run in a conduit. There is only one available grounding electrode, although the plan does cause for the conductor to be extended and connected to cold water pipes as well as building steel. What he showed on the site plan is one ground rod in a concrete encasement, with 1"PVC-#2 ran to several points where its connected to the chain post using a ground clamp. The closest of these points to the building is then shown as a home run to CWP and structural steel as mentioned. It may be easier to define things per artcile 250: is the conductor from the ground clamp on the post to the ground rod a GEC? I may not be thinking straight as I just got home after yet another one of those days (aka every day!) but I'll try to dig deeper into it later tonight. I know it's a tricky article but I'm a bit more disappointed with myself every time I have to go back and look at an article that I've read a hundred times already...

 

[George Stolz]

I may join my befuddled colleagues, but are you asking how the NEC is requiring you to size a conductor that the NEC doesn't require? :huh:

 

[skeshesh]

I may join my befuddled colleagues, but are you asking how the NEC is requiring you to size a conductor that the NEC doesn't require? :huh:

Certainly not. I did not bring up the NEC until the very end when I was trying to think about it in a different (wrong) way just to wrap my head around the problem. I was looking for a method of calculation hence the post under this section not the NEC.

I think I arrived at the solution browsing around IEEE 80 and NEC 780. Since this situation is not one around a substation and it's a low lightning risk the only way that fence could become energized is from the electrical 'stuff' around it. All the small pumps there are fed from a 100A or 225A 120V/208Y 10K panel way downstream in the distribution. I dug up an older reference good ol Jim Dunger posted ( http://forums.mikeholt.com/archive/index.php/t-80411.html) and used the formula from pages 535-536 or so. For a rough approach I used a 0.5s fault duration and 10K fault and the appropriate conductor constants arriving at about 83kcmil which is around #1.