On May 3, 11:38 pm, Tony Hwang wrote:
I experienced a direct lightning strike on a 7 story building. In the
basement there was a large(I mean LARGE) scale data center which I was
in charge of.
The strike clobbered all the data stored in mass storage sub system
requiring 3 days' total system restore. I think when surge is BIG,
nothing can be protected from it.

Broadcasting electronics atop the Empire State Building and World
Trade Center were struck 25 and 40 times annually without damage.
Commercial broadcasters with antennas thousands of feet up also suffer
such strikes and cannot suffer damage. Your telco with switching
centers in every town; with their $multi-million switching computer
connected to overhead wires all over town; must suffer such surges
routinely without damage. Mid 1900 research indicates a thunderstorm
typically creates maybe 100 surges - and no damage.

Likely the outgoing path through that scale was via concrete floor.
What was the incoming path? Well what in that circuit was damaged?
Or was it only data loss, which means hardware protected itself when
too much surge current was permitted inside the building?

How many days did your telco require to reprogram that switching
center computer after every thunderstorm? They can suffer 100 surges
during every thunderstorm and not even suffer data loss - let alone
hardware damage? Exactly. Effective protection means every wire in
every incoming cable has a short connection to earth via a 'whole
house' protector AND separation of up to 50 meters between the
protector and electronics. Not used are plug-in protectors. Any
protection that would work at the equipment is already inside the
equipment. Not acceptable is damage from lightning – even data loss.
And if damage does occur, telco located and corrected an earthing
defect.

Another example in Central Florida where Orange County's emergency
response system suffered lightning damage. Lightning damage
eliminated by fixing the defect - earthing:
http://www.psihq.com/AllCopper.htm

Same is described by van Deursen and van der Laan when lightning
caused damage to a nuclear hardened maritime radio station. Did they
cry, "Woe is me. Nothing can stop lightning damage"? Of course not.
Their IEEE paper describes how earthing defects (human failures) were
fixed. Lightning damage directly traceable to a defect in the
earthing system – human failure.

It is routine to suffer even the most massive surges and no surge
damage. Lightning routinely strikes communication facilities on Hoher
Peissenberg mountain in southern Germany - without damage.
Researchers even mounted electronics equipment to measure the currents
of each surge. Did direct lightning strike destroy that electronics
and communication equipment? Or course not. It is routine to suffer
direct strikes without electronics damage. However the human must
first learn what provides that protection - especially proper
connections to earth ground.

What makes surge protection so challenging? We can test other
designs. But we cannot test the surge protection system. Therefore,
when damage does occur, the responsible human locates and learns his
mistake - often must correct a defective in that earthing system.
Numerous professional citations also describe learning from the damage
because lightning damage is so easily avoid.

A protector is only as effective as its earth ground. As Phil
correctly notes:
But it is a matter of how much you want to spend on it.

Simple earthing to meet NEC requirements creates significant
protection. High reliability facilities may spend even hundreds more
to obtain but a little more protection. On average, a destructive
surge may occur once every seven years. How much would you spend.
$20 for some earthing rods to significantly upgrade protection; or
$hundreds to also have protection installed in Central Florida:
http://members.aol.com/gfretwell/ufer.jpg
How much was that data worth? A question asked of others since Tony
Hwang routinely denies this stuff. He suffered massive station
damage. Then he declared nothing can protect from lightning even
though his industry peers says completely otherwise.