Lighting Arrestor Safety Zone

A very good answer! I live very near a tall steel TV antenna and have seen lightning strike well inside the radius.

GA for a better than GREAT answer!

How people can finance a building and not have enough left over to do the final job of Lightning conductors properly is really laughable........if it was not so serious.

We have a saying in Germany "He saves money no matter what it costs!". I wonder if the original Blogger will understand that fully?????

Very GA. Unfortunately I can only give a GA.

I have seen several diagrams showing a "Cone of Protection" for lightning arrestors, but you gave an excellent discussion on how and why they can "break down" and not work as expected. Lightning is unpredictable stuff, and the severe storms that produces lightning should be respected and treated with extreme caution. Any safety system for these types of events will not be perfect, so plan accordingly.

Years ago I had to do some Lightning protection design for an outdoor PA system for the USAF in Colorado Springs, Colorado, USA. I ran into the concept of "cone of protection", and found that, as stated before, it was really a statistical probability distribution plan, and thus, unless you wanted to gamble with life, not reliable. I also ran into the idea of the "charge drain" method of protection, which essentially designs for a method to "bleed off" the charge between the earth and the clouds above before a strike path could form, and a strike occur. NASA, along with the NOAA had done a great deal of research on both, and had come to the tentative conclusion that the "charge bleed" idea had more merit than the "cone of protection idea", but that neither was a worthwhile gamble with lives at stake. The result of my studies was that we devised a "sacrificial charge gap", very similar to the carbon block protector that used to be installed in telephone lines to house, but with the addition of Metal Oxide Varistors, used to attentuate the charge, but more importantly to slow the rate of current change, and create a 90 degree lag between the voltage change leading edge and the current change leading edge. This, all together, had the effect of "snubbing" the total power buildup to keep either E or I relatively low compared to each other at any point in the waveform. Keep in mind, though, that relatively low is indeed relative, when the research at the time put the average highest current passed in a lightning strike in excess of 275 KA, and the maximum voltage in excess of 375KV, with the entire process of ramp up to maximum and decay to zero, even for a strike with multiple pulses, taking no more than 10 ms. So there is a potential for E*I, if neither is snubbed, and the development of both current and voltage is near simultaneous (there is some delay due to cloud capacitance no matter what the storm conditions are, thus I will always lead E by a small phase shift) to reach the product of these two values, or around 103 trillion watts of power (someone else will have to convert that to Joules, but its definitely in the GJoule range, I'd think) in 10 thousands of a second. And that's one BIG ARC! The result, modelled, but never directly testable, was a total power maximum in the low MW range, as opposed to the high GW range. And when we experienced our first lightning storm after installation, the two 3USD (in 1000 quantities, in 1986) devices in the charge path blew out and saved our 8 mile long, 000 gauge copper cable runs, and our 3.75 KWatts of power amps, at the expense of 3 speakers costing a total of 45USD, and two "protectors" at 6USD. We were satisfied, to say the least.

Incidentally, the "multiple stroke path" discussed before is called a "stepped leader" and is the formation of a plasma channel of superheated air, which has significantly lower resistance than the same distance in even damp air (on the order of 1 ohm per 100 feet, vice 10Kohms per inch). It doesn't last long, and as noted is susceptible to strong air currents, being often bent as much as 30-90 feet sideways, over its typical 5000-15000 foot length, in the winds commonly experienced in a lightning storm. In fact, strange to note, when NOAA flew a retired F-104 Starfighter (I think that's the proper name, the number is right, according the articles I read) through a storm to attract a lightning bolt, wanting to study that channel bending phenomenon, they found small "rosettes" where the arc hit the outer fuselage of the plane. But what was amazing was that the entry point, during a supersonic pass through the storm, was near the tail section of the plane, and the exit point was nearer the nose. Since that model is mostly just a fire-filled tube with a pilot sitting on it, the arc had to enter the skin, travel forward by some path through the engine, and exit on the opposite side of that fire-filled tube. The assumption was that the current travelled around the skin of the engine, bypassing the supersonic, superheated jet of gases propelling it. Examination of the tube after the engine cooled showed that the arc in fact entered aft, crossed into the gas channel, travelled upstream against the flow, entered the opposite wall of the tube, passed through it, and left the plane. It did all of this with a right-angled diversion AGAINST the flow inside the engine. The plane was also caught on video tape "dragging" the lightning bolt as it passed through the plane, so that the bolt was bent almost 500 feet out of its path, to form a 90 degree vee between its top and bottom ends.

I know all that has nothing to do with the idea of "safe zones" and "cones of protection", but it appears to me that if we know so little about lightning after all the studies we've done that we can neither explain nor predict that behaviour, we're a lot better off ignoring statistical probabilities and designing for all the built-in safety we can get.

And with potential Giga-Watt discharges at stake, even that will still be plenty iffy.

Micah

One point not mentioned. Inductance per linear foot of conductor is higher for conductors with small perimeters. Ben Franklin found this out when one of his early installations melted the down lead. After that he used hollow tubes or rectangular cross sectioned iron conductors.

Low inductance is important with a fast current rise time to minimize V = L * dI/dt. If V gets too high, cross flashing results, as mentioned by other posters. Low resistance is important, too, to minimize V = R * I. Atmospheric charge bleeds rapidly through the lowest available impedance path.

HAVE FUN! BE SAFE!

Good point well explained. Thanks.