Surge Protective Devices and High Ra.

well the Ra /Ze /Zs are not really what you need here - for the leading edge of short transients there could be as much current flowing into the capacitance of the wiring acting as one plate and the rest of the universe as the other, as there will be actually flowing into or out of the electrode(s).

As a first cut rule of thumb, consider the electrode as some resistance to terra-firma, in series with a micro-henry or so per metre of wire to get to the electrode frome where you are measuring,  and the whole lot in shunt with perhaps a few hundreds of pF or nanofarads depending on the size of the installation, between every wire and the ground beneath it.
(To do it properly involves looking at the wires over ground as interconnected and coupled transmission lines and solving for velocity factor and characteristic impedance. I suspect most folk prefer the rule of thumb.)

At 50Hz the resistance dominates, and all the OSG approximtions are good, but for short duration events, that is no true and the L-C transmission line effects can no longer be neglected as small.

If the electrode is 'far away' compared to the time of the transient then even if it had a fantastically  low impedance no current would flow, as the electrons would not get there in time.

Having said that, the L, N and E will all bounce together once the MOV and gas discharge have entered their conducting state. This stops sensitive electronics 'doing the splits' being between the lines.That is all the SPD guarantees. What voltage the lines are at relative to something far away is not controlled - they may all bounce up to a few kV relative to another earth electrode if you could bring the voltage of one in from far away without bonding it.

Note also that for the same reasons we too are not affected by short pulses in the way that we are by steady voltage - the 10kV kick of an electric fence may exercise your 'vocabulary of opportunity', but is nothing like as deadly as the same voltage  (and current) would be persisting for a cycle or two of 50Hz instead of being all over in a hundred microseconds or so.

Mike.

well the Ra /Ze /Zs are not really what you need here - for the leading edge of short transients there could be as much current flowing into the capacitance of the wiring acting as one plate and the rest of the universe as the other, as there will be actually flowing into or out of the electrode(s).

As a first cut rule of thumb, consider the electrode as some resistance to terra-firma, in series with a micro-henry or so per metre of wire to get to the electrode frome where you are measuring,  and the whole lot in shunt with perhaps a few hundreds of pF or nanofarads depending on the size of the installation, between every wire and the ground beneath it.
(To do it properly involves looking at the wires over ground as interconnected and coupled transmission lines and solving for velocity factor and characteristic impedance. I suspect most folk prefer the rule of thumb.)

At 50Hz the resistance dominates, and all the OSG approximtions are good, but for short duration events, that is no true and the L-C transmission line effects can no longer be neglected as small.

If the electrode is 'far away' compared to the time of the transient then even if it had a fantastically  low impedance no current would flow, as the electrons would not get there in time.

Having said that, the L, N and E will all bounce together once the MOV and gas discharge have entered their conducting state. This stops sensitive electronics 'doing the splits' being between the lines.That is all the SPD guarantees. What voltage the lines are at relative to something far away is not controlled - they may all bounce up to a few kV relative to another earth electrode if you could bring the voltage of one in from far away without bonding it.

Note also that for the same reasons we too are not affected by short pulses in the way that we are by steady voltage - the 10kV kick of an electric fence may exercise your 'vocabulary of opportunity', but is nothing like as deadly as the same voltage  (and current) would be persisting for a cycle or two of 50Hz instead of being all over in a hundred microseconds or so.

Mike.