How lightning rods and protection systems really work

01

Does a lightning rod attract or repel lightning?[1][2]

Neither, in the way most people mean. A rod does not pull down strikes that would otherwise miss, and it does not bleed enough charge off a storm to stop one. What it does is control where a strike that was already coming attaches, and where the current goes afterwards.

The distinction matters because both myths lead to bad decisions. If rods attracted lightning, adding one would raise your risk, and people would rationally leave them off. If rods drained clouds, a single rod would protect a whole neighbourhood. Neither is how the physics works.

A descending leader spends most of its journey indifferent to what is below it. Only in the last tens of meters does the ground start to matter, when the field at pointed objects rises enough to launch upward streamers. Whichever streamer connects first decides the attachment point. A rod wins that competition on purpose, because it is higher, sharper and better grounded than the roof around it.

The charge-drainage idea has been tested directly. Arrays of sharp points sold as strike-eliminators have been placed at instrumented sites and struck anyway. The current a corona point can bleed is a few hundred microamperes. A thundercloud regenerates its charge thousands of times faster than that.

02

What is the rolling sphere method?[1]

It is the standard way to decide where rods go. The designer imagines a sphere of fixed radius rolled over the structure from every direction. Wherever the sphere touches, lightning can reach, so that point needs protection. Space that the sphere cannot reach is inside the protected zone.

NFPA 780 uses a 150 foot (46 m) sphere, which corresponds to a striking distance for a peak current near 35 kA. The radius is not arbitrary: a leader carrying more current makes its final jump from further out, so a bigger sphere models a stronger strike and a smaller one models a weaker strike.

That is also the method's built-in limitation, and the standard says so. Strikes weaker than the design current have a shorter striking distance than 46 m, so a smaller sphere would fit into gaps the design sphere rolls straight over. Those low-current strikes can slip past the rods and attach inside the supposedly protected zone. Protection systems reduce risk to a defined level rather than eliminating it.

Rolling the sphere is why tall structures get rods along every roof edge and corner rather than one spike at the peak. Edges and corners are exactly where a sphere makes contact first.

Air terminals

the rods themselves, placed where the rolling sphere touches

Down conductors

at least two paths to ground, so current divides instead of concentrating

Grounding electrodes

where the current actually leaves the system and enters soil

Bonding

ties metal pipework and structure to the system so current cannot arc sideways to reach it

03

Why does bonding matter more than the rod?

Because lightning current does not politely stay in the conductor you gave it. If a nearby metal pipe sits at a very different potential during the strike, current jumps the gap to reach it. Bonding ties everything to the same system so there is nothing left to jump to.

During a strike, a down conductor's potential rises by a large amount relative to anything not connected to it. That difference is what drives a side flash, an arc that leaves the intended path and crosses to plumbing, wiring or structural steel. Side flashes are a common cause of lightning fires in buildings that had rods installed.

This is the same reason indoor lightning safety advice singles out plumbing and corded electronics. Those are the paths current takes once it is inside a structure, whether it entered through a protection system, the service line or the ground itself.

04

Does a lightning protection system protect the electronics inside?

No, and this is the most common misunderstanding about them. A structural system handles the direct current path and keeps the building from burning. It does nothing about the voltage transient induced on wiring inside. Electronics need surge protection, which is a separate installation.

The two systems solve different problems at different energy scales. Tens of thousands of amperes going around your house is a structural fire question. A few thousand volts appearing briefly on a data cable is an equipment question. A building can have flawless rods and still lose every device plugged into a wall.

Insurers and standards treat them as complementary for that reason, and a full installation normally includes both.