This possibility is obvious from the technical documents I read, and it was what I worried (and I'm still worrying!) about. Thanks for the comment, I now know i

This possibility is obvious from the technical documents I read, and it was what I worried (and I'm still worrying!) about. Thanks for the comment, I now know it's a real danger, not just a theoretical one. Because of this, I eventually abandoned the mesh network experiment. I don't want to blow the school up and pay for it, and potentially have someone killed, just for a faster Intranet. And I decided that if I'm going to do the experiment in the future, I'll just setup a gateway in an open space, somewhere middle but away from the buildings (so the gateway can pick the signals from those buildings), and run the system on solar power. In this case, lightning protection becomes easier - just ground the system and it should be fine.