Lightning Protection Question

[Supersteele21]

I'm having a little trouble with an engineer (at least I hope so) on a lightning protection installed on a hospital building I am finishing up.
We installed PVC from the ground level up to just below the roof (inside the columns). From there, we used RMC nipples (approximately 3-4' in length) to penetrate through the roof. The lightning protection system is bonded at every nipple through the roof, and bonded again to the ground loop at ground level. I am catching some flack from the engineer who says that "NFPA 780 and probably UL require that conduit be bonded at both ends of the system".

I think that the RMC being bonded to the lightning protection system as well as the ground loop on the Earth side is sufficient. What he wants us to do is take out all the RMC and install PVC nipples- which will be a tremendous headache for all involved: new cad welds, roof patching, etc. Especially with the state health inspection scheduled for September 13th!

I am a new member here, but I have often used the forum as a research tool and certainly appreciate anyone's help/advice.

Thanks!

 

[GoldDigger]

Because of the inductance of a conductor passing through a ferrous tube of any sort and any length, and the high frequency (fast rise time) nature of lightning related currents, the NEC requires that a Ground Electrode Conductor (GEC) must be bonded at each end to any ferrous conduit or sheath that it passes through. This is independent of any other bonding on either side of the ferrous metal tube. This allows the current to flow in the tubing itself instead, avoiding the iron core inductor effect.
I am sure the same considerations apply to a lightning protection system down conductor.
Depending on the wire and the fittings on either end of the nipples you may be able to bond compliantly at both ends of each nipple. If you cannot, I am afraid you will indeed have to replace them with PVC, brass, stainless or other non magnetic tubing.

Sent from my XT1585 using Tapatalk

 

[Dennis Alwon]

If there is a section that requires the bonding on both ends then it seems like that is what you have to do. It seems ridiculous that it would need 2 bonds. I would like to see the section he mentioned. Perhaps on a long run it would make sense but not on a 4' piece, IMO

 

[GoldDigger]

If there is a section that requires the bonding on both ends then it seems like that is what you have to do. It seems ridiculous that it would need 2 bonds. I would like to see the section he mentioned. Perhaps on a long run it would make sense but not on a 4' piece, IMO

In theory, even a thin ferrous washer, let alone a short nipple, will cause a measurable and possibly significant inductance.
A good rule of thumb is that if you have a situation where you would have to worry about inductive heating you also have to worry about voltage drop.

Sent from my XT1585 using Tapatalk

 

[sameguy]

Ground clamp.

 

[Adamjamma]

Still learning but, If he grounds the nipple at the PVC joint to an egc, doesn't that solve it? If it is already grounded at other end?

However, most lightning systems I dealt with were basically just 1/0 or 2 guage wire run from the roof to the ground rods... so maybe I am not understanding at all what is going on here.

 

[Ragin Cajun]

It the GEC is bare, bonding is not required since it's in continuous contact the length of the conduit.

 

[Supersteele21]

The real issue here is that the PVC from ground level is continuous, converting over to RMC before it goes through the roof, where it is bonded to the Lightning protection loop. All wire has been pulled and cad welded. The issue being that of the process the wire, pulling it out, cutting a section of PVC out to make room for a choker (bonding connector) on the bottom side of the RMC nipple, and then re-pulling and cad welding all of the drops (which there are a substantial # in a 70,000 square foot building) will be quite timely and expensive. Obviously, code does not care about time and expense- but I was hoping that the engineer's request was just excessive, and not required.

 

[jumper]

It the GEC is bare, bonding is not required since it's in continuous contact the length of the conduit.

Code reference please?

 

[tom baker]

It the GEC is bare, bonding is not required since it's in continuous contact the length of the conduit.

What code section allows this?

 

[LarryFine]

Perhaps on a long run it would make sense but not on a 4' piece, IMO

Even a 6" metallic pipe must be bonded on both ends. Lightning can cause induction that inhibits the wire's conductivity.

It's not for conductive bonding, but for inductive bonding. More like not running a single phase in metallic conduit.

 

[winnie]

I am with GoldDigger on the theory about why this is an issue, so the question is one of 'what _exactly_ does NFPA 780 require?'

Does the bond need to be done with something 'rated for lightning protection system' or is something like a split grounding bushing acceptable?

Do you have options other than removing everything and re-doing?

For NEC required grounding (but I don't know about NFPA 780) you have available 'split' grounding bushings which can be attached to the conduit with the wire in place, and you also have various grounding clamps available. 'Lay in lugs' are also available for connecting to wire without having to cut it.

-Jon

 

[LarryFine]

Yes, you can use a split grounding bushing, and even a jumper to connect the bushing to the wire.

 

[jaggedben]

Yes, you can use a split grounding bushing, and even a jumper to connect the bushing to the wire.

I'm imagining the rigid nipples threaded into female PCV connectors and seeing the difficulty he has. Nowhere to install that bushing.

I'm assuming these are the lightning down conductors and not grounding electrode conductors. In which case, anybody got a copy of 780? NEC requirements for GECs are irrelevant I think.

 

[jumper]

I'm imagining the rigid nipples threaded into female PCV connectors and seeing the difficulty he has. Nowhere to install that bushing.

I'm assuming these are the lightning down conductors and not grounding electrode conductors. In which case, anybody got a copy of 780? NEC requirements for GECs are irrelevant I think.

Same, 4.4.2

 

[Supersteele21]

All-
Jumper is exactly right in his description of the system with the female adapter changed to a RIGID coupling. I do have a workaround that the UL inspector said would be acceptable. Our RIGID nipples are attached to the metal decking of the roof with a blank 4x11 cover and lock rings on either side of the decking. The inspector said that would suffice for the bonding, because of all of the other bonding (roof & ground loop/building steel). Maybe he's letting us slide a little, but I consider this a lesson learned.

 

[sameguy]

All-
Jumper is exactly right in his description of the system with the female adapter changed to a RIGID coupling. I do have a workaround that the UL inspector said would be acceptable. Our RIGID nipples are attached to the metal decking of the roof with a blank 4x11 cover and lock rings on either side of the decking. The inspector said that would suffice for the bonding, because of all of the other bonding (roof & ground loop/building steel). Maybe he's letting us slide a little, but I consider this a lesson learned.

Sounds pretty weak to me, if bond to building steel is ok you could still use waterpipe ground clamps. Cut the pvc out and split bolt. Sounds like someone is trying to work with you, but a 4sqr. Cover ? Trek screwed to the roof deck? No.