There are indeed assumptions - one small lighting circuit on a board fed by a hundred amp submain, you can indeed assume the source impedance is negligible compared to that of the final circuit. However circuits that supply one or two big loads maybe not.

Another way to think about that is how the design voltage drop has been partitioned, as that is also in effect the conductor resistances.

Actually measuring of very low impedance loops - without dedicated test rigs that uses representative fault currents anyway, becomes a bit fruit machine once you get into the many kA end of things, and can be dominated by meter leads and the cleanliness of connections.

So what to do ?

Well it depends how well you need to know, and what for - usually we are ensuring that the fuse or breaker operates promptly for a fault at far point, where a rounding up resistanace and down in current is needed, but the other case is to check that  the let through energy is not so high that cables will be damaged with a fault near the origin, when the safe assumption is to round the current up and resistance down.

Not all texts are clear for which of those two situation the assumptions are appropriate.
In reality it all needs to be taken into account, but taking into account may be to decide that ' it's not the dominant uncertainty, I'll ignore it for now'

Mike

Some calculations I have come across simply calculate Zs from the combined resistance of the cables within the installation

They shouldn't. The basic formula is Zs = Ze + R1 + R2 - where R1+R2 are the "internal" resistances, but Ze is the "external" (i.e. source plus impedance of the public supply wiring).

Faults inside equipment are trickier - windings can add very considerably to the loop impedance and often the installation installer can have no knowledge of actual numbers (e.g. when designing to supply a socket outlet into which anything could be plugged). Generally BS 7671 designs cover the installation and then it's down to appliance standards to work out how to provide protection for internal faults. In general though, the larger the internal resistance, the lower any resulting touch voltages, so longer disconnection times become tolerable, so often it not quite as bad as it might appear at first sight.

  - Andy.

simonturner83 said:

Some calculations I have come across simply calculate Zs from the combined resistance of the cables within the installation, it strikes me that this omits the impedance associated with the source (e.g., a supply transformer) and load (e.g., an AC motor) devices, where the latter should have some influence in the case of a fault to earth in the motor windings.

That approach does not conform to BS 7671, which requires you to take into account "worst-case" conditions.

You are correct, that the source impedance Ze or Zdb is also required ... and also that the (R1+R2) used should also be calculated for 70 degrees C (this is the purpose of the 0.8 factor rule of thumb if comparing "as-measured" values with Tables 41.2 to 41.5 of BS 7671) ... or a higher value of temperature for cases where conductor temperatures under load are higher (not recommended unless you can prove that terminals of equipment connected to a 90 or 110 degree conductor can withstand that temperature).