Testing if supplementary bonding is required.

Actually what we should assume for the "worse case"  scenarios are not as clear cut as you imagine - This recent example on this forum  demonstrates what can happen with a defective RCD, that is not even on the property where folk are getting the shocks. This one is seriously complicated by  having a lower electrode impedance at the fault than at the substation, and is an example of this kind of thing that the regs do not consider as  it is 'unlikely' . But occasionally the unexpected happens.


It is also faults like this that make me very careful in descriptions to distinguish terra-firma earth voltage  (the voltage with your feet in muddy stuff outdoors) from "earth" that comes via a wire, as supplied via a CPC and the DNO neutral.


Mean while I'd agree, anything with no wiring to it and reading more than a few tens of k ohms to terra-firma is floating, and need not be bonded, as it cannot sensibly carry enough current to give you a bad shock.

Personally I'd not cut it as fine as a limit of 23k, as the quality of the connections to things like bits of  pipe set in  concrete floors and so forth are a humidity (and voltage) dependant thing, and you may be measuring on a dry day. But even if you make the "action limit" value  say 50K, then you still do not need to bond shelf brackets and toothbrush holders and all the other things that connect to the wall or floor with a few rawlplugs.

Which is good, cos we do not, and if we did, we would get laughed out. (shadows of the 15th edition, one of the least followed versions of the rgs..)

Mike

Actually what we should assume for the "worse case"  scenarios are not as clear cut as you imagine - This recent example on this forum  demonstrates what can happen with a defective RCD, that is not even on the property where folk are getting the shocks. This one is seriously complicated by  having a lower electrode impedance at the fault than at the substation, and is an example of this kind of thing that the regs do not consider as  it is 'unlikely' . But occasionally the unexpected happens.


It is also faults like this that make me very careful in descriptions to distinguish terra-firma earth voltage  (the voltage with your feet in muddy stuff outdoors) from "earth" that comes via a wire, as supplied via a CPC and the DNO neutral.


Mean while I'd agree, anything with no wiring to it and reading more than a few tens of k ohms to terra-firma is floating, and need not be bonded, as it cannot sensibly carry enough current to give you a bad shock.

Personally I'd not cut it as fine as a limit of 23k, as the quality of the connections to things like bits of  pipe set in  concrete floors and so forth are a humidity (and voltage) dependant thing, and you may be measuring on a dry day. But even if you make the "action limit" value  say 50K, then you still do not need to bond shelf brackets and toothbrush holders and all the other things that connect to the wall or floor with a few rawlplugs.

Which is good, cos we do not, and if we did, we would get laughed out. (shadows of the 15th edition, one of the least followed versions of the rgs..)

Mike