[ghshephard]

We have a lot of our equipment tested in Florida (where they have some large testing centers) for lightning tolerance. I've been told, that you don't actually verify the equipment can be safely struck by lightning - as anything that is hit by lightning, is immediately destroyed. The objective is to ensure that equipment nearby a lightning strike doesn't suffer secondary damage (possibly from a power surge in the line). We install lightning arrestors near our electronics, but anything that actually gets hit is replaced.

Airplanes though, are supposedly struck by lightning, and continue to work. I'm guessing that this has something do do with them being suspended in air, and not having a connection to ground.

[Animats]

Antenna towers get hit all the time, and the transmitters and receivers survive if installed properly. First, there's a spark gap between the antenna and ground, with big metal contacts, copper or silver, a short distance apart. That diverts most of the lightning bolt to ground. Then the feed line for the antenna has a big inductor, a coil made of heavy busbar, usually in a grounded can. This is often placed through the wall of a grounded metal equipment enclosure.

The inductance blocks a fast risetime lightning bolt, forcing the energy to the spark gap. A few hundred volts will still get through that. So following that there's a gas tube protector, which is essentially a neon tube which will ionize and short to ground. (Phone lines also have those at the central office end.) Following that is a MOV, as in a surge suppressor, to dump the remaining surge into ground. What's left after than can be tolerated by most RF electronics intended for such applications.

If this didn't work, radio wouldn't work in Florida. It's not that this stuff is expensive compared to the equipment it protects. It's that the front end stuff is big; #4 copper cables, big spark gap units, heavy ground rods, and solid metal equipment enclosures with welded seams.

Somewhere right now, a cellular tower is taking a lightning hit and restarting itself without damage.

[ghshephard]

You seem to know a lot about this, so all I can do is reiterate that when we take our electronics to florida to be tested, our QA guys make it clear that no company's electronics are designed to survive a lightning strike - including ours. (Though we are designed to resist a 22 calibre bullet). We install a lot of lightning arrestors, mandatory in Florida, São Paulo, Singapore -- and we seem to do fine in those areas (knock on wood) - but I've been told that's to prevent a nearby lightning strike from running down the antenna line into our electronics. If our RF gear (which is mounted outdoors in weather rated containers) gets hit with a direct strike - 100% guaranteed destruction.

[eigenvector]

I think you and GP are actually agreeing. The goal of a lightning protection system is to ensure that when a structure (antenna, wind turbine, transmission tower, etc.) is hit by lightning, the lightning energy is provided with a low impedance path to earth such that any sensitive equipment inside or attached to the structure is not exposed to currents and voltages beyond its design limits.

There are other considerations as well for personnel safety, like ensuring that the ensuing ground potential rise doesn't expose someone standing near the earthing point to a shock hazard.

Going back to the original comment, lightning protection in wind turbines is more or less a solved problem. Direct hits on blades are routine and any utility-class wind turbine will be designed with an appropriate system for the area in which it is located. Now, when your turbine is on a big rock, that's a problem and you may have to blast and backfill to achieve a sufficiently low earth resistance.

[classicsnoot]

From all i have read, direct strikes on gear is a grab bag of destruction. I've read that in the Vietnam Conflict guys would get shot and the bullets would bounce off of their skin [strange effect of velocity, turbulence, and chance], so i bet there are examples of "look mom, no destruction" but i tend to side with you, at least as the guy managing this project: at no point am i going to say "cannot fail", "works everytime" or any version of the two. If you have never been to Maine, it is beautiful. If you have never built in Maine, don't get attached. The sea wind destroys everything. I kept my tools covered every day on a 42 day stint except 1, and the rust is ubiquitous. So i build to fail. The real question is, how long can i draw out the failure?