[BC] lightning & grounding

[Clive Warner]

>When they are used, Alan's point about grounding and conduction from
>ground to the device is obviously VERY important. A very low impedance
>between device and ground is essential if they are to carry significant
>"dissipation" currents.

** I decided to check this out. First, I noted that in the most amusing
'massive strike' story, the earthing strips were only 0.032" thick! Thirty
thou copper is fine for handling RF but totally inadequate for proper
grounding, and the installer should have known better. In other words that
tower was improperly and inadequately grounded. Lucky that the strike did
not go through the rebar in the foundation and explode the base of the
tower.
The correct way to do it is to use one-eighth strap, not wimpy 0.032 stuff.

> This suggest effective use may be problematic
>with series fed AM towers, and also suggests a need for extremely good,
>very low resistance bonding for grounded towers, as well as a very good
>ground termination (IOW a pattern of long rods, and possibly chemical
>augmentation).

** This last is the only really sensible and verifiable contribution to the
argument that I have read, to date.
     From the EMC Handbook, Vol 3 (Don White & Associates):-
"A building ground may exhibit an impedance to earth of the order of 0.5
ohms"
     (seems about right; I have always tried to achieve <0.3 ohm in my own
installs)
"... a typical strike from a charged-cloud source, having a potential to
ground of about 100 MV, results in a current of about 30 KA. Assuming a
negligible impedance of the lightning arrester and structure, the neutral
GROUND (their emphasis) point is suddenly elevated to 15KV (30KA x 0.5ohm)

I suggest this is the reason why damage can still occur when grounded masts
are struck. The base of the mast, and therefore the coaxial shield, is
momentarily elevated to 15KV. However the building is relatively remote and
so its ground voltage is 15KV different from the 'ground' voltage at the
mast.

A cage fed mast will exhibit a maximum of 800V difference between the feed
point and the coaxial screen, (see my earlier post for the calculations). I
consulted a senior antenna engineer at BAE Systems, who confirmed that the
folded-umbrella is the most lightning-resistant structure of any type.

With ground radials the ground impedance is extremely low, making the
grounded AM radiator the most lightning-proof structure. Of course FM and TV
masts have no radials and so it behooves the engineer to install an adequate
earthing system. My old friend the late Martin Newton used to rely on deep
pits filled with truck radiators scavenged from nearby scrapyards, and urine
(more on this later, Barry!).

Many engineers drastically underestimate what is required to provide an
effective ground.
Assuming soil resistivity of 50 ohm-cm, achieving a ground resistance of 0.5
ohms requires either:

a) 10, 3/4 rods, evenly spaced over a 7,500 sq ft area and driven to depths
of 30 ft.
or
b) 7, 3/4 rods, evenly spaced over a 12,000 sq ft area and driven to depths
of 30 ft.
(From EMC Handbook)

Incidentally, I truly hope that US RF sites are exempt from the NEC. See
sections 250-82 and 250-83 - the code specifies that a water pipe shall
'always be used as the grounding electrode' reinforcing my belief that the
US NEC is a load of complete BS.

And finally, some earlier messages read, at least to me, as if sharp points
act to discourage a stroke. I read on P 9.4 of the EMC Handbook:

"When the ionizes column of lightning strikes a building, it will seek out
higher-elevation sharp surfaces and low-impedance paths to earth."

- Clive Warner