On Jul 13, 1:52*pm, Owen Duffy wrote:
Owen Duffy wrote :

...

But, firstly, you should determine if there are regulatory
requirements, such as NEC etc.

Is "NFPA 780: Standard for the Installation of Lightning Protection
Systems" a relevant standard in your jurisdiction?

Owen

Owen
It is relevant as a consensus standard but it is not adopted as local
or State law. Do you have a link to a copy that can be read online?
--
Tom Horne, W3TDH

Tom Horne wrote in
:

....
Is "NFPA 780: Standard for the Installation of Lightning Protection
Systems" a relevant standard in your jurisdiction?

Owen

Owen
It is relevant as a consensus standard but it is not adopted as local
or State law. Do you have a link to a copy that can be read online?

Ok, well the question is whether you give it importance, or stay with NEC.
I think it turns out that you have a copy, read it and make you own mind
up.

If it was me, I wouldn't waste the money on an inadequate protection
scheme.

Owen

Tom Horne wrote:
On Jul 13, 1:52 pm, Owen Duffy wrote:
Owen Duffy wrote :

...

But, firstly, you should determine if there are regulatory
requirements, such as NEC etc.
Is "NFPA 780: Standard for the Installation of Lightning Protection
Systems" a relevant standard in your jurisdiction?

Owen

Owen
It is relevant as a consensus standard but it is not adopted as local
or State law. Do you have a link to a copy that can be read online?


NFPA 780, like NFPA 70, is a copyrighted document *sold* by NFPA.
However, there *are* online copies of various provenance and age around.
http://www.atmo.arizona.edu/students...A_780_2004.pdf

Unfortunately, the bare code doesn't tell you much about the "why" for
various code provisions, so if you're thinking of going "off code" for
one reason or another, you don't have a lot of information to tell you
whether it's a good idea.

There's also some interesting seeming inconsistencies.. NFPA 780
requires a minimum length of a ground rod of 8 feet (4.13.2.1) but also
requires that they extend vertically not less than 10 feet into the
earth (4.13.2.3(A))) The figure makes it clear.. the top of an 8 foot
rod is 2 feet below the surface of the soil.

NFPA 780 says 29 square millimeters for main conductors (6 mm in
diameter or a strip that is 1.3mm thick x 22.3 mm wide).. That's AWG 6
roughly.


There's also a great site by Carl Malamud: publicresource.org that has
all the California Building Codes (including an older rev of the NEC)
although it doesn't have NFPA 780 on it, as far as I know.

On Jul 13, 2:18*am, Owen Duffy wrote:
Tom Horne wrote in news:e802f6fa-b0e1-471b-bf31-
:

Can anyone make a recommendation, based on actual training and
experience, as to what width and thickness of copper strap would be ...

In this part of the world, we have an Australian / New Zealand Standard
(our version if you like of ANSI, BS etc) which explains the rationale
behind lightning protection, a method of estimating the downcurrent for
protection design purposes and a process for designing down conductors.

Broadly, the scheme is that downconductors are designed to withstand a
few donwstrokes in quick succession without melting the down conductor.
If you work from a peak current of 20kA, it would lead to a down
conductor in copper of at least 25mm^2 which is about #2 to you folk.

I regularly see hams recommend much thinner down conductors, and can only
assume that there is not regulatory guidance or requirement, and I wonder
at the effectiveness of using #6 as often recommended, especially
aluminium as is often the case. Note that reducing conductor size is a
double whammy, you increase the resistance (so the power), and decrease
the mass that has to be heated to melting point, and so the energy
required.

But, firstly, you should determine if there are regulatory requirements,
such as NEC etc.

The question of equipotential bonding conductors ought be dealt with in
the same way, though that is not to imply that they will be the same
size.

Owen

Owen
Believe it or not the NEC only calls for number ten American Wire Gage
(AWG) or 5.261 (mm)2 for protective grounding conductors. Bonding
conductors between electrodes are only required to be number six AWG
or 13.30 (mm)2. So leaving aside the bad joke that is the NEC
requirements I'm trying to get some idea of what best practice might
be.
--
Tom Horne, W3TDH

Tom Horne wrote:

Owen
Believe it or not the NEC only calls for number ten American Wire Gage
(AWG) or 5.261 (mm)2 for protective grounding conductors. Bonding
conductors between electrodes are only required to be number six AWG
or 13.30 (mm)2. So leaving aside the bad joke that is the NEC
requirements I'm trying to get some idea of what best practice might
be.


Tom is a bit confused here about the purpose of NEC vs NFPA 780..

The bonding requirements in the NEC are designed to keep the building
from burning down in the event of an accidental fault to an energized
conductor. The basic requirement is that it carry enough fault current
for long enough to trip the overcurrent protection device on the
energized conductor.

It's not for lightning protection, per se. (although NEC bonding will,
incidentally, provide some degree of protection against induced transients)

I'd also note that AWG 10 wire is more than sufficient to carry a 50kA
pulse for the 20 microseconds or so that a lightning stroke lasts
without melting.
Using the Onderdonk equation, you can calculate that a AWG16 copper wire
will carry about 90kA for a 20 microsecond pulse. AWG10 should be able
to carry 4 times that much. AWG6, 10 times, because it scales with cross
sectional area.

Having only really paid attention to this recently, I noticed that in
Rome (a place with a fair number of thunderstorms), they use fairly
small down conductors (AWG 10 or 6, just by eye), and similar for 7
story tall wooden pagodas in Nara, Japan (another place with lots of
thunderstorms).

I'm not quite sure where the fashion for 2/0 grounding conductors comes
from (maybe Phelps-Dodge has a representative on the NFPA 780 review
committee?grin)

On Jul 13, 2:18*am, Owen Duffy wrote:
Tom Horne wrote in news:e802f6fa-b0e1-471b-bf31-
:

Can anyone make a recommendation, based on actual training and
experience, as to what width and thickness of copper strap would be ...

In this part of the world, we have an Australian / New Zealand Standard
(our version if you like of ANSI, BS etc) which explains the rationale
behind lightning protection, a method of estimating the downcurrent for
protection design purposes and a process for designing down conductors.

Broadly, the scheme is that downconductors are designed to withstand a
few donwstrokes in quick succession without melting the down conductor.
If you work from a peak current of 20kA, it would lead to a down
conductor in copper of at least 25mm^2 which is about #2 to you folk.

I regularly see hams recommend much thinner down conductors, and can only
assume that there is not regulatory guidance or requirement, and I wonder
at the effectiveness of using #6 as often recommended, especially
aluminium as is often the case. Note that reducing conductor size is a
double whammy, you increase the resistance (so the power), and decrease
the mass that has to be heated to melting point, and so the energy
required.

But, firstly, you should determine if there are regulatory requirements,
such as NEC etc.

The question of equipotential bonding conductors ought be dealt with in
the same way, though that is not to imply that they will be the same
size.

Owen

Owen
The NEC only requires 5.261 (mm)2 for the protective down conductor
and 13.30 (mm)2 for the bonding conductor between electrodes. Since
those sizes are at best a bad joke I was hoping to elicit best
practice advise on what size the conductors should actually be as well
as advise on how to accomplish the bonding of the interior single
point grounding buss bar to the exterior grounding conductors and
Grounding Electrode System.
--
Tom Horne, W3TDH

On Jul 15, 9:18*pm, Tom Horne wrote:


Owen
The NEC only requires 5.261 (mm)2 for the protective down conductor
and 13.30 (mm)2 for the bonding conductor between electrodes. *Since
those sizes are at best a bad joke I was hoping to elicit best
practice advise on what size the conductors should actually be as well
as advise on how to accomplish the bonding of the interior single
point grounding buss bar to the exterior grounding conductors and
Grounding Electrode System.
--
Tom Horne, W3TDH

It's not that bad a joke.. If the ground connection is good, #10
is plenty thick enough. In fact, it would barely get warm if it
took a strike. Of course, if the connection to ground is bad,
it will be toast. But so would a lot of heavier gauges also..
The connection to ground is the critical factor in such a case.
But I would still follow what the local code says.
The main reason I'm making this post is only to clarify that
under proper conditions, #10 is plenty thick enough to safely
route the strike to ground with no damage to the wire.

Tom Horne wrote:


Owen
The NEC only requires 5.261 (mm)2 for the protective down conductor
and 13.30 (mm)2 for the bonding conductor between electrodes. Since
those sizes are at best a bad joke

Perhaps you could explain why you think it's a bad joke? Do you think a
13 square mm conductor couldn't carry the strike current? (it can)

Or, perhaps, you're thinking that there are some other design criteria
that might push one towards a larger conductor (mechanical strength in
the face of icing and storms might be one).

Jim Higgins wrote:
On Mon, 19 Jul 2010 09:23:53 -0700, Jim Lux
wrote:

Tom Horne wrote:

Owen
The NEC only requires 5.261 (mm)2 for the protective down conductor
and 13.30 (mm)2 for the bonding conductor between electrodes. Since
those sizes are at best a bad joke
Perhaps you could explain why you think it's a bad joke? Do you think a
13 square mm conductor couldn't carry the strike current? (it can)

Or, perhaps, you're thinking that there are some other design criteria
that might push one towards a larger conductor (mechanical strength in
the face of icing and storms might be one).

Maybe E=IR has something to do with wanting a larger conductor. The
voltage between the strike point and true ground is going to be the 20
- 100 kA of the strike times the resistance of the down conductor from
the strike point to true ground. With a smaller conductor,
fewer/shorter ground rods, or other conditions that raise the
resistance of the path to ground that voltage will be higher and if
high enough the strike will seek additional paths to ground by arcing
to nearby objects closer to ground potential.


Resistance isn't actually a big deal here. It's all about inductance on
that microsecond rise time pulse. And there's not much difference in
inductance between a AWG 6 and 2/0 (it's very weakly dependent on
diameter and strongly dependent on length.. 1 microhenry/meter is a good
estimate, pretty much independent of diameter)

The other problem is that for fast transients, skin effect means that
the AC resistance goes more as the diameter than as the cross sectional
area (hollow tubes work just as well as solid conductors).

So, the net effect is that you need to design for several things:
1) the wire not melting..
2) The wire not breaking from mechanical impact (ladders hitting it,
lawnmowers, etc.
3) The wire not breaking under electromagnetic forces (this is why you
don't want loops and why NFPA 780 says 8" bend radius.. while a 1
microsecond pulse at 10kA won't melt a AWG 10 wire, if it's in a loop,
it will destroy it from EM forces)

You'll see heavier conductors where they have to be able to move (say on
a gate or actuated device), not only for mechanical life, but also
because the flexible wire is more subject to destruction by EM forces.

Side flash is a consideration, but usually accommodated by making sure
your downleads are far from potential victim circuits.

Jim Lux wrote in
:
....
The other problem is that for fast transients, skin effect means that
the AC resistance goes more as the diameter than as the cross
sectional area (hollow tubes work just as well as solid conductors).

The problem is that while we might characterise the raw excitation caused
by lightning, and use assumptions about the shape, rise and fall times
and peak field strength, the response of circuits (such as those that
include the down conductor) is quite different, and it is unsafe to
assume in the general case that skin effect is fully effective for all or
even most of the energy spectrum.

Perhaps that is why some of these standards tend to treat the conductor
as having a resistance equal to that implied by just the conductivity (or
resistivity) and CSA. It might be conservative, but then standards tend
to be so.

Having seen the results of fairly detailed EM modelling of EMP and
lightning excitation of major infrastructure, and the effects of some
small changes to the model, I wonder a bit about the effectiveness of
some measures... but over engineering probably saves the day in a lot of
cases.

The real danger with lightning protection is that a half baked approach
my give the implementor some comfort, but actually increase the risk of
adverse outcome.

The most thorough and consistent practice I have seen is that employed
here in mobile phone base stations. Sure, they are occasionally damaged
by lightning, but the vast majority of lightning incidents do not cause
permanent damage.

Owen