Dave, thank you for that excellent and informative reply! I really appreciate
you taking the time to write it.
73,
Dave AB7E
------Original Mail------
From: "K1TTT" <K1TTT@ARRL.NET>
To: "'Towertalk e-Goups'" <towertalk@contesting.com>
Sent: Sun, 09 Aug 2009 12:13:33 +0000
Subject: Re: [TowerTalk] "Faraday Shield" for Coax and Control Lines
Note, I play the part of an amateur on here, but in real life I write
simulation software for designing HV power line lightning protection systems
and have participated in several high voltage and high current experiments
at the lab where I work. This has included high current impulse testing of
full size ground rods and radials. We have an impulse generator that can
create about 25KA or 5MV pulses with several microsecond durations for
simulating lightning.
Some simple rules for induction:
Currents are induced in loops. The bigger the loop the more current. The
closer together the conductors are the smaller the loop and the less current
will be induced IN THE LOOP.
Observations about lightning impulses and ground:
Because of the fast rise time and finite propagation velocity of currents in
the ground the farther apart ground connections are, the higher the voltage
difference between them.
NESC rule:
As I understand the rules, ALL facility grounds must be tied together with a
minimum size conductor back to the service entrance. Maybe someone can
qualify this, but I don't 'think' there is a distance limit. This is a
power frequency safety issue, NOT a lightning safety design issue.
Lightning physics:
When lightning strikes the ground or an object connected to the ground it is
not 'dissipated'. There is always an opposite charge on the ground that has
been attracted by the charge in the cloud that the lightning is being
attracted towards. When the downward leader and the upward stream meet the
charges cancel each other starting at the connection point and propagating
up the lightning channel and downward and outward along the ground at a
finite velocity (approx .3c).
For perspective:
A conductor in the ground has a velocity factor of about .3. This gives a
speed of about 100m/usec. So in a 1usec rise time of a typical lightning
stroke hitting a tower, the wavefront can travel 50m to the shack and back
and still affect the voltage on the tower. One way to interpret this is
that in the 1usec rise time at 50m the voltage will just start to rise at
the shack end when it is near the peak at the tower. This has little
application for amateur designs, but helps scale the problem as far as speed
and distance.
How these apply to the topic at hand:
When properly installed the coax shield is connected to ground at the tower
and at the SPG. While this doesn't legally qualify as an NESC ground
connection, it IS connecting those two grounds. If you don't parallel the
coax shield with a REAL conductor it WILL try to carry some portion of the
lightning current away from a struck tower.
This does NOT depend on the distance to the shack. Because of the fast rise
time there is not enough time for the wave to propagate to the shack end of
the cable and back within the rise time... and even if it did that would
make it worse since when the reflection returns it would INCREASE the
current in the coax shield because of the ground on the shack end. So short
coax would carry more current, but during the critical rise time even long
coax could get significant currents depending on how many other paths to
ground there are at the tower.
Conclusion 1: YES, put a large conductor between the tower ground and the
shack ground, no matter how far away it is.
Next, on the separation of the ground... there are 2 important cases here.
First, the near miss case:
As I stated above from simple physics, the bigger the loop the higher the
induced current. For conductors above ground the air filled loop is easy to
see. The current induced in the loop created by raised cable runs and the
ground is relatively easy to calculate and can be used to predict the peak
voltages at gaps or arresters in the loop. Simple rules are: the higher the
cable is above ground the higher the voltage/current, and the longer the
loop the higher the voltage/current... so grounding long cable runs at
intermediate support posts would reduce the voltages seen at the ends from
near misses.
In the case where there is a 'ground' conductor on the pole with the current
carrying conductors it is still the area of the loop that is the critical
factor. The closer the ground conductor is to the current carrying ones the
smaller the induced voltages will be. This is independent of it being above
or below the other conductors... though there is the secondary effect of
voltages induced on the ground conductor itself that makes it harder to
evaluate.
For buried conductors the near miss calculation is not so easy because the
cables are in a partially conductive media that is also a dielectric. And
the 1-3' range normally used by amateurs to bury them is usually well within
the skin depth of normal soil. But generally because of the surrounding
soil the currents induced in loops in the ground are greatly reduced from
what they would be in the air. So while there may be some small advantage
to keeping the conductors close to each other, from an induced voltage
standpoint it is not as critical. My instinct on this would be to ignore it
for coax shields since they should be connected to ground at both ends
anyway. It would be different for low voltage signal conductors like rotor
cables. I would not count on the ground attenuation to protect them so
arresters should be used at both ends.
For the direct stroke to a tower:
This is more problematic because the currents are much higher. In this case
I would worry most about the initial rising edge where the currents and
voltages are the worst.
There are 3 things I would consider here:
1. In the design of high voltage transmission lines it is well known that
you improve the protection of the current carrying conductors by keeping the
overhead shield wire close to them. This is because the current flowing in
the shield wire induces a voltage in the nearby power conductor that reduces
the voltage difference between it and the tower. This is easily calculated
and can result in 10's of KV's changes in voltages over separations of many
meters. While this effect may also be reduced in a buried situation I would
expect it to still be significant because of the high currents and closer
physical proximity. For a ground buried next to the cables we are talking
inches instead of many meters... on the other hand hopefully only a small
percentage of the lightning current is present in the buried ground as
opposed to the high percentages seen in high voltage line overhead shield
wires.
2. Electric fields from the buried ground in the soil would also help
prevent punctures of the outer coax jacket. Because the buried ground is
conducting some of the lightning current it is changing the electric field
in its vicinity. This should be propagating at about the same speed as the
wavefront on the coax shield and would result in reducing the voltage
difference across the outer insulation. If the coax were buried separately
the full voltage difference between the shield and the soil would be higher.
3. If 2 conductors are buried in different trenches between a tower and SPG
there is a strong likelihood that the propagation velocity may be slightly
different in each of them because of depth, soil conditions, vicinity of
other conductors, etc... It is also likely that they may not be as near the
same length as they might be in the same trench. While this may be a small
effect in amateur sized installations, our voltages are much lower than
utility lines also so the difference in arrival time of the two wavefronts
at the far end of the conductors may be significant.
Another possibility to consider, though likely a smaller probability event
because of the relatively short length of buried amateur cables and the
proximity of towers and buildings... it is well known that lightning strokes
to the ground can penetrated several meters through the soil to strike
buried power lines, pipelines, etc. There is even a name for the fused soil
channels left in the ground after these events.(Fulgerites sp?) In these
cases having a good ground conductor above, or at least near, the signal
cables 'may' provide some protection by intercepting the penetrating stroke.
The combined effects of the above make it hard to recommend separating the
ground in a different trench from the coax/control cables. There is an
advantage to be gained from the induced currents coupling together the
various cables, and the construction is easier to start with... and with the
possible bad effects of propagation differences and the added cost it just
doesn't make sense.
David Robbins K1TTT
e-mail: mailto:k1ttt@arrl.net
web: http://www.k1ttt.net
AR-Cluster node: 145.69MHz or telnet://dxc.k1ttt.net
> -----Original Message-----
> From: Dave - AB7E [mailto:xdavid@cis-broadband.com]
> Sent: Sunday, August 09, 2009 07:07
> To: Towertalk e-Goups
> Subject: Re: [TowerTalk] "Faraday Shield" for Coax and Control Lines
>
>
>
> Hhmmm ... the answers I've received here so far don't necessarily combine
> to a clear conclusion, possibly because some people think I'm talking
> about conducted currents from a strike on the tower. I'm not ... I think
> the tower is reasonably well grounded at the base and I'm more interested
> in the chance for induced currents on the runs to the shack due to a
> nearby lightning burst (not necessarily a direct hit).
>
> First off, here's why I think the tower is well grounded, as a couple of
> folks have questioned. It's a heavy duty freestanding tower with the
> three legs embedded about five feet in a 9' by 9' by 6' foundation. The
> foundation is loaded with LOTS of rebar (see the details at www.ab7e.com)
> all wired together, and each piece of rebar in the entire bottom grid is
> cadwelded together using #4 solid copper wire. The bottom of the embedded
> tower legs are cadwelded to the bottom grid of the rebar cage with one #4
> solid copper wire for each leg. The bottom of each tower leg is also
> cadwelded to its own ground rod buried a few inches under the foundation
> with #4 solid copper wire (no, I don't buy into the exploding foundation
> myth). Lastly, each tower leg has two #4 solid ground wires (six wires
> total) running about thirty feet away from the tower (every 60 degrees
> radially) with cadwelded ground rods spaced every few feet long them.
>
> The shields of the coax, Heliax, etc are bonded to the tower both at the
> top of the tower and at the base. The control lines currently have surge
> suppressors at the top of the tower, and soon will have additional ones
> both at the base and at the shack. The coax and Heliax have grounded
> differential mode surge suppressors at the shack. All feedlines are DC
> shorted either at the antenna or the shack (the ICE suppressors have a
> shunt coil), or both.
>
> So, to reiterate, my concern was mainly about induced common mode currents
> on the feedline shields or control lines, and I have a hard time
> understanding why induced currents on the ground wire from the tower to
> the SPG would in turn induce greater currents on the feedline shields or
> control wires than would already be induced on them of their own accord
> from the field energy associated with the lightning. So why run the
> ground wire in a separate trench as so many have advised? Doesn't make
> sense to me unless the concern was about CONDUCTED ground currents from
> the tower.
>
> I also still do not understand why having a ground wire above the
> feedlines would not provide some measure of shielding from INDUCED
> currents if that ground wire was connected to ground rods every few feet
> along it's length. The concern about induced common mode currents is
> mostly with respect to ground, is it not? Possibly I am wrong on that
> assumption. And if such a grounded shield wire doesn't perform a useful
> function, why do the electric utilities use it on telephone poles? The
> answer isn't critical to my installation one way or the other, but I'd
> like to understand.
>
> 73,
> Dave AB7E
>
>
>
> ------Original Mail------
> From: "Dave - AB7E" <xdavid@cis-broadband.com>
> To: "Towertalk e-Goups" <towertalk@contesting.com>
> Sent: Sat, 08 Aug 2009 12:40:09 -0700
> Subject: [TowerTalk] "Faraday Shield" for Coax and Control Lines
>
>
> I asked this question over a year ago and didn't get any responses, so I'm
> going to try it again.
>
> The trench from my shack to my tower is roughly 180 feet long, and since
> it was dug with a backhoe it is a couple of feet deep. The ground wire to
> the SPG, the Heliax, and the various control lines will all be in that
> same trench. The tower itself is extremely well grounded.
>
> I live in an area with the potential for violent summer lightning storms,
> and I'm wondering if the ground wire to the SPG near the shack might
> provide a shielding effect to minimize induced common mode surges on the
> Heliax and control lines, much like the top wire on electrical utility
> lines is used. I can position the ground wire a foot or so above the
> other lines and Cadweld it to ground rods spaced regularly (every ten
> feet?) along the run.
>
> Anyone have a thought on whether or not that would provide any extra
> protection from induced surges?
>
> 73,
> Dave AB7E
>
>
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|