"it is recommended that an additional connection to earth, by means of an earth electrode ..............."

Yes?

That might be useful for preventing normal earth leakage currents raising the installation's metalwork to a hazardous voltage if the means of Earthing the installation is lost entirely (e.g. break in PE, but not N), but I can't see it being a sensible mitigation to a broken PEN in general. Generally you'd need to achieve something like  Ra * I < 50V - where I is the installation's load - to have any sensible effect - so fine where the installation draws less than an amp (single street lights for example) - but even for a modest domestic installation Ra would have to implausibly low to have any sensible effect.

   - Andy.

"it is recommended that an additional connection to earth, by means of an earth electrode ..............."

Yes?

That might be useful for preventing normal earth leakage currents raising the installation's metalwork to a hazardous voltage if the means of Earthing the installation is lost entirely (e.g. break in PE, but not N), but I can't see it being a sensible mitigation to a broken PEN in general. Generally you'd need to achieve something like  Ra * I < 50V - where I is the installation's load - to have any sensible effect - so fine where the installation draws less than an amp (single street lights for example) - but even for a modest domestic installation Ra would have to implausibly low to have any sensible effect.

   - Andy.

AJJewsbury said:

Ra * I < 50V

But if a number of installations downstream of the break have an earth electrode, then R_a(effective) is the parallel resistances of Ra for each installation (assuming no other earthing of the PEN conductor, and no extraneous-conductive-parts).

The good thing about that, is we often hear that there could be high numbers of properties downstream of the break. The higher the number of properties affected, the greater the population at risk ... yet if all had electrodes (and/or extraneous-conductive-parts), the lower R_a(effective) ...

Now, you could also argue that I goes up with the number of installations ... BUT it's also the more likely the break is in a three-phase portion of the distribution network, so if there are n properties downstream of the break, it's only n/3 times as much current (not n times), and also the overall earth electrode resistance requirement is quite possibly lower (depending on the maximum phase imbalance downstream of the break).