Lightning protection.

[toddro1971]

We are installing a Nextell receiver antenna and amplifier on top of our plant. It will be mounted on a 10 foot stick of 11/2" EMT on the tallest point of the tallest building. Do we need lightning protection, the building is a one story steel building with tin siding about 40 feet high. The antenna is about 15 feet tall all together. Right now we have zero lightning protection except for ground rods at the panels. Can we bond to these? Do we need a dedicated system? Located in central Florida.

 

[bphgravity]

Re: Lightning protection.

Yes you do, but not the kind you are probably thinking. Forget the traditional air terminal to down conductor to ground conductor arrangement. This will not provide you the protection you are looking for. I would invest in TVSS or other surge arresting devices.

 

[toddro1971]

Re: Lightning protection.

Thats what I was thinking BPH, my concern is we are essentially putting a big areal on top of the building. I am concerned that it will eventually get hit by lightning. Ok, if it fries the antenna, no big deal. But where will this voltage dissipate to. We have a server room directly below the antenna. But if we make an effort to ground the antenna we are creating an even better lightning rod. Danged if we do or dont. Should we try to isolate the antenna from the building and protect the amplifier with a TVSS. Or go ahead and ground the antenna with a seperate ground rod and down conductor?

 

[ron]

Re: Lightning protection.

All the grounding electrodes, down conductors, miscellaneous grounds, must be bonded together.

There is no national code that requires lightning protection, but the recommended practices of NFPA 780 would probably lead to a recommendation of installing protection (air terminals, down conductors, TVSS, etc).

 

[bphgravity]

Re: Lightning protection.

Most lightning protection literature will tell you that a lightning protection system incorporating an air terminal as the strike point does not induce or attract a lightning strike. However, it is known that ground based objects emit upwards streamers that connect to downward leaders from passing clouds, thus creating the short and flash event.

So it is a rock and a hard place situation. It depends on what you think is more probable. This is theory and not law. If lightning were to strike the antenna / mast structure on the roof, current will flow over all the conductove parts of the system. The more resistive the path, the more likely damage will occur. Therefore, providing a low impedance path to earth via a down conductor connected to a ground rod or rods may reduce the surge on other paths to a less damaging value. Then again, it may not.

I am in the process of gathering data and establishing an experiment of my own with lightning protection and a lack there of. I am hoping to build a few structures in an open field and then see if I can get lightning to strike one and destroy everything in it.

 

[dereckbc]

Re: Lightning protection.

toddro, this is quite a complicated subject which is to detailed to go into here on the forum, but I suggest you do what common carriers and cell providers do in regard to antennas.

I design protective grounding systems for data, telephone, and cell sites. A good place to start is Polyphaser at http://www.polyphaser.com/ppc_ptd_home.aspx
There you will find documents and products that address exactly what you are talking about.

Basically the ideal situation would be to have a lightning protection system with air terminals, down-conductors, and ground ring. Antenna mask and coax shield are bonded to the roof conductors. Run the coax/power straight down to about ground level and install a ground bar where the coax enters the building, and install the antenna discharge unit and TVSS on the ground bar which is bonded directly to the ring.

The catch here is none of these measures other than those outlined in Article 810/820 are required by the NEC. Everything I speak of is well beyond the requirement and scope of the NEC and is a design issue.

[ November 21, 2005, 11:10 PM: Message edited by: dereckbc ]