[vhfpmr (OP)]

If you look closely at this building you will see that the lightning conductor runs horizontally in a ring around the base of the roof just above the gutter. There are numerous strips coming down to it from the top of the roof, but none connecting it to the ground.

Eh?

[Bored chemist]

Maybe the ground connection is round the back.

[vhfpmr (OP)]

I can't see a good reason for making the circuit longer than it needs to be though. It makes sense to continue the vertical strips all the way to the ground in one.

[Halc]

Looks like plumbing to me, maybe some mechanism to combat ice dams.
Perhaps it's conduit for electrical heating since I don't see how water would work..
I lived in a farm house with such defense, but it was a zig-zag pattern that included a run through the gutters.

[alancalverd]

Earth bonding strip for a future solar panel array?

[Eternal Student]

Hi.

    I can't see too clearly on the picture but it looks like there is a lot of old fahsioned iron guttering to collect rain water off the roof along with old-fashioned heavy iron down-pipes from the gutters. If the lightning conductor (assuming that is what it is) is connected to that then you're almost done.   Just add a few good grounding points on the waste pipes at ground level.   Many buildings did / do still have their main ground connection achieved through a water pipe.

Best Wishes.

[Bored chemist]

I can't see a good reason for making the circuit longer than it needs to be though. It makes sense to continue the vertical strips all the way to the ground in one.

Because you need to protect the whole building.
The problem I see is the right angle bends.

And the magic of the internet lets me find someone who explains why.
https://ch00ftech.com/2011/12/15/why-you-dont-make-right-angle-traces-and-why-lightning-rods-are-pointy/

[vhfpmr (OP)]

Because you need to protect the whole building.

How is a building with a loop at gutter level that's grounded remotely from the lightning conductor spikes better protected than it would be with a strip along each of the ridge lines grounded immediately beneath each conductor spike?

[Petrochemicals]

My guess is it is a metal frame roof. It looks like a large span, so some sort of metal 'A' frame. There is at least 1 ground, to the left of the window by the sign.

https://www.buildingenclosureonline.com/ext/resources/ARWWCA/Winter-18/Lightning-Protection/Photo-Aug-19-11-34-52-AM-WEB.jpg

[vhfpmr (OP)]

A metal frame under a tiled roof on a building erected 90 odd years ago? I don't think so.
It's a traditional construction, brick, with tiles on wooden rafters. The span's only one classrom plus one corridor wide.

[Petrochemicals]

A metal frame under a tiled roof on a building erected 90 odd years ago? I don't think so.
It's a traditional construction, brick, with tiles on wooden rafters. The span's only one classrom plus one corridor wide.

https://en.m.wikipedia.org/wiki/Steel_frame#History

[vhfpmr (OP)]

Yes, I know, but why would it be a steel frame, it's not a high-rise, and it's not a large span, it's just a traditional building like the vast majority of the others built in that era. And equally importantly, if a steel frame's protecting the building why put a lightning conductor on it.

[evan_au]

old-fashioned heavy iron down-pipes from the gutters

"Skin effect" means that high-frequency components (like lightning impulses) only travel through a thin skin on the outside of the conductor.
- That is why lightning down-conductors are made of a braid of many fine wires (with a large surface area) instead of one big conductor.
- Iron has poor conductivity
- I expect the magnetic interaction with iron would also impair the ability to handle high-frequency components...
https://en.wikipedia.org/wiki/Skin_effect

Horizontal Lightning Conductors

There is another instance where you get horizontal lightning conduction, with Fulgurites on a beach.
- The lightning turns the sand into a glassy tube
- When the lightning reaches the more conductive water, it sometimes turns into a branching pattern at the level of the water table...
https://en.wikipedia.org/wiki/Fulgurite

[Bored chemist]

That is why lightning down-conductors are made of a braid of many fine wires

The ones I have seen have been metal strips (again, that reduces inductance).

[Eternal Student]

Hi.

   "Skin effect" - I didn't know that, thank you.   
"Lightning impulse" -  I'll guess it's not genuine AC with Voltage swings from +V to -V many times.   I'll assume it's a single DC spike,    from  0 (volts or amps)  to +V_max   (or +A_max ) and back to 0  over a short time interval.    I've never really looked at a Lightning strike and seen it as a plot of  Voltage (or Current)  vs.   time,   maybe it has mutiple sinusoidal components with different frequencies.       I'll have a passing interest if anyone can point to a plot of how the voltage or current from a lightning strike changes with time.

Best Wishes.

[Bored chemist]

Every time I try to measure the current/ voltage vs time plot for lightning two things happen.
The oscilloscope evaporates noisily and the villagers surround the castle carrying pitchforks.

But the "official" version used for testing looks like the curve in this paper.
https://eprints.soton.ac.uk/267056/1/UHV_paper_Long.pdf

[paul cotter]

Hi ES, I can't give you the data on lightning which you seek( I had a great reference book covering this subject but it was lost in a business liquidation ), however I wanted to expand on the skin effect. A very high di/dt as occurs in lightning will produce the skin effect with a pure dc discharge as the intense magnetic field pushes the conduction electrons to the surface. An alternative view is that a fourier analysis of such a spike will yield a lot of high frequency components.

[vhfpmr (OP)]

"Skin effect" means that high-frequency components (like lightning impulses) only travel through a thin skin on the outside of the conductor.
- That is why lightning down-conductors are made of a braid of many fine wires (with a large surface area) instead of one big conductor.

All the lightning conductors I ever recall seeing are flat copper strips. Braiding it won't make any significant difference to the skin effect because the strands aren't insulated from each other, which is why radio engineers use Litz wire to reduce skin effect. (At MF/HF, above that the usual ploy is to use silver plated copper.)

Digressing a bit, a novice mistake for radio engineers is to use tinned copper wire, in which most of the current flows through the lossy plating instead of the copper, or bare copper wire, which has the same effect by the time it's oxidised. A common technique for reducing the cost of UHF feeder is to use copper plated steel for the inner conductor, because the current isn't flowing in the steel.

"Lightning impulse" -  I'll guess it's not genuine AC with Voltage swings from +V to -V many times.   I'll assume it's a single DC spike,    from  0 (volts or amps)  to +Vmax   (or +Amax ) and back to 0  over a short time interval.    I've never really looked at a Lightning strike and seen it as a plot of  Voltage (or Current)  vs.   time,   maybe it has mutiple sinusoidal components with different frequencies.

The bit highlighted is correct.

If you have a repetitive pulse waveform with a period T1, a pulse width T2, and a rise time T3, then it will have a spectrum with components spaced apart by a frequency equal to 1/T1. At frequencies which are an integer multiple of 1/T2 there will be a null in the envelope of the spectrum where the amplitude reduces to zero, and the highest frequency component will be at 1/T3.

From this is follows that if you have a single pulse, T1 is infinite, and the spectrum will be continuous, without discretely separate frequencies, and with nulls every 1/T2, up to a maximum of 1/T3. (If you had a single infinitely narrow impulse with zero rise time it would have a spectrum which is a horizontal line flat to infinity.)  In the case of the pulse in BC's reference, the rise time is ~1.2uS, so the highest frequency component you'd expect to see in the spectrum would be about 800kHz, but since the spectrum is continuous, the majority of the energy will be at much lower frequencies. (More than half the energy will be below 400kHz)

Skin depth in copper is:
Freq.        Skin depth (μm)
50 Hz   9220
60 Hz   8420
10 kHz   652
100 kHz   206
1 MHz   65.2

[evan_au]

Lightning consists of one or a few closely-spaced DC impulses. A Fourier analysis shows that many frequencies make up this DC impulse.

I have done lightning testing on telecommunications equipment with a test generator that could generate several waveforms that corresponded to various standards, with peak voltages around 10,000 Volts or so. As I recall, two of them were DC impulses like the graph shown by BC (with rapid rise and slower fall, but different time constants). A third had an AC component, which you might get if the surge were inductively/capacitively coupled from an adjacent wire; in this case you don't get such a strong DC component, the higher frequencies are enhanced compared to the lower frequencies, and you get some "ringing" from the inductive/capacitive resonance.

The most spectacular example of lightning testing was at a visit to a national lightning test facility, which could generate impulses over a million volts. Understandably, the test engineer checked everything over very carefully before he took his hands out of his pockets!

[Bored chemist]

I read that there are about 44 lightning strikes per second and I guess that they have been doing that for millions of years.

But I bet that not a single one of them has ever precisely followed the graph in that paper I cited.

One class of thing that gets hit fairly often is a transmission antenna.
And they are designed to resonate, so I suspect they have really weird looking V/T graphs.