Put effort into lightning protection for exposed items. There are nets and other forms of hail protection available. I guess they get deployed by climbing up there right before a storm turns nasty. Not everyone would want to do that. There are also specified resistant panels, and a handy hail-risk map What You Need To Know About Solar Power Hail Damage
This is what I have in mind: Two non-conducting posts or insulators along the apex of the roof Pointy lightning rods, at each end of the ground cable For esthetics, dull gray or black wire How to keep birds from resting on wire? Grounded away from the grid, service lines (on left) feeding the house Pass through the bedroom attic to outside and down Substantial ground rod to earth A lightning strike will still induce a substantial EMF in the regular house wiring. But TBD if surge protectors would work on Neutral, Hot_1 and Hot_2 to the earth ground. Fortunately the breaker panel is in the bedroom closet on the right Bob Wilson
Late thought, the cable to ground will be run inside: A lightning strike will induce a significant magnetic field which is the start of a huge, EMI pulse. By running the ground wire through a ferrous conduit, to the extent capable, it should reduce the magnetic pulse. How much is question to test. Any retired electrical engineers out there? Found this research paper: https://www.osti.gov/servlets/purl/1875497 Bob Wilson
Would you just get a young person up a ladder to remove those dead leaves? I tire of holding back commenting on that. The horizontal wire with be a bird perch no matter what. Why are you so intent on denying that? Oh you Canute you. But I can't see why you need it. The two grounded vertical pointy sticks will attract lightning better than metal frames of PV panels. Oh here's a thought - if PV frames have sharp edges, spend an hour or two on the ground softening the edges and corners. Like with a fingernail file. Very soft aluminiumiumium.
I have a question for @jdenenberg about running an insulated, heavy gauge ground cable/wire inside a ferrous conduit. Florida lightning research shows that as long as both conduit and cable are grounded at the same earth point, no problem. My first thought was the induced magnetic field would be somewhat attenuated by the ferrous, outer tube. This would reduce the magnetic field inducing an electric pulse in the house wiring. Then I thought, "rail gun." The high currents should generate a radial force that might make the ferrous tube into a lightning powered, pipe bomb. In effect converting a lightning fire hazard into something worse. But I may be over thinking the problem. The grid service line comes in on the Northwest wall of the house but runs in the crawl space to the circuit breaker panel at the opposite side of the house. Individual circuits are in the crawl space with the exception of the utility room on the North side extending into the back yard. If I run the lightning ground cable 90 degrees to the house wiring into the back yard, Northeast side of house, any induced EMF would significantly be weaker than if the grounding wire ran parallel: 200 A service line (red) comes from left to Northwest facing wall. Service passed through crawl space to the circuit breaker (red box) panel in the Southeast bedroom. All but one circuit pass through the crawl space. Power to utility room, Northern most extension, passes through attic Solar panel wiring in conduit in attic 13 solar panels on Southwest facing roof 3 solar panels on Southeast facing roof over porch Lighning canary line about a meter above apex of tallest roof Earth ground on Northeast back yard. Tallest point, actual a lightning rod on the canary line at the earth ground. As for the birds, they would draw attention to the lightning canary line and leave guano behind. For rain collected, garden water not a problem. For other purposes, I'm not a fan of bird guano. There will be an owl puppet on one end and each panel array will have a rubber snake fixed to the bottom. The goal is to minimize nesting below the panels. Bob Wilson
Home solar panels have some maintenance: cleaning critter collections under the panels So I'm going to need to go up from time to time and I want to inspect some of shingles, about 4 years old. My home owner's insurance wants to know the 'damage resistance' which I had never heard of before. So this is how I've left the ladder for tomorrow: Should there be a ladder problem, the roof of the BMW is a much shorter distance than to the ground. The ladder base is also held by a 2,800 lbs BMW bumper: The major risk is transitioning from roof to ladder but I think it is covered. I will wait for my GF to help 'supervise' and call 911 if I screw up. But I'll be the one going to the roof since I'll also be inspecting shingles and later, the panels. Finally, installing the lightning protection and bird repellant owl and rubber snakes. Bob Wilson
Running the lightning ground through the attic seems like a bad idea, besides adding a potential leak point in the roof. There is an insane amount of energy in a lightning bolt. Forget EMF, and consider heat running along uninsulated wood. The bolt can also jump from the cable. More potential points it could jump to inside than running along the outside. Lightning rod installs just run along the roof and exterior walls. Looks like there is heavy gauge copper flashing for the grounding cable. Understanding The Basics - Residential Lightning Protection Systems | 4LP No need for a suspended wire. You do want another ground point coming off the far aerial instead of relying on all of a strike traveling across the roof.
My current thinking is running the grounding wire under the exterior, eve of the roof. I want it to keep as dry as possible. Bob Wilson
You really should get the correct stand-off arms for the ladder. They attach near the top and will prevent the ladder from crushing the shingles at the edge of the roof like your ladder is doing. Those shingles are likely going to crack at the bend and fall off in the future. The arms also provide a very steady perch and make it easier to get on and off. And remember 4 to 1 for height/distance from base. Too steep and you're much more likely to tip backwards or slide sideways. All that extra height you have on that ladder may make it even harder to get on and off. Slide that thing a foot or so to the right and you have a pretty good step off!
As far as I can tell, that paper is looking at effects where lightning current is sharing normal building electrical wiring, and wires are insulated and sharing a conduit. Your situation is much closer to that in the NEC section on "Enclosures for Grounding Electrode Conductors" (which see). I believe there's a similar provision (maybe even more emphatic) in the codes that are specific to lightning-system conductors, but I'm less familiar with those. The gist of the NEC provision is that the electrode conductor is allowed to run in a ferrous enclosure only if the ferrous enclosure itself is made an electrically parallel current path. It must either be an electrically continuous enclosure all the way from where the electrode conductor starts to the grounding electrode it leads to (and connected to both), or, if not, then every segment of ferrous enclosure that the conductor passes through must be bonded to the conductor itself at both the entry and the exit end. That may defeat, or partly defeat, your purpose to minimize the magnetic field seen outside, as instead of having the ferrous material enclose something carrying a current, it ends up carrying a big share of the current itself. So why do the NEC folks insist on this? Inductance. A lightning current has a crazy fast rise time, so even tiny inductances in the ground path matter. Every place your electrode makes even a ¼ turn around a corner, well, that's another ¼ turn, adding to the inductance. Count those and minimize. Could a different routing need an even smaller fraction of a turn? The larger the inductance, the more huge the voltage spike seen at the upstream end for such large dI/dt. It can be huge enough for some lightning to give up on your conductor and follow something else. A ferrous enclosure around the conductor—when it isn't bonded and sharing the current—is effectively running the conductor through an inductive choke. That's why the NEC guys say nay. https://www.ecmag.com/magazine/articles/article-detail/codes-standards-minimizing-choke-effect-understanding-ferrous-metal-raceway-bonding
