Well… first of all doing some very quick studying in magnetic losses… and secondly referring to 543.6.1 and making perhaps a sensible deduction… where a protective conductor is not incorporated in the same wiring system as the live conductors or in their immediate proximity or separated by ferrous material, this may cause magnetic losses which could affect disconnection times in much the same way as circuit length affects disconnection times.

is that a close enough for rock and roll just of things here?

Pleased to see that the query is being answered.

More generally, another potential resource is the Legal Deposit libraries (previously, copyright libraries). Your county library service may also hold copies.

It may seem odd, but British Standards On Line (a wonderful resource!) does not include BS 7671, but presumably that is because it is not the sole author.

Threading a current through a block of magnetic material raises the inductance of the wire, for sure,and if the 'mu'  value and the geometry of the surrounding magnetic material , in this case a cylinder, is known, then the inductance per inch of conduit or whatever can be easily deduced as the one turn limit of the toroid core problem. 

But note that the magnetic properties of conduit steel are not well specified, and any  mu value in the range from maybe  500 (nut and bolt steel) up to perhaps 30,000 for magnetically optimized steels is possible, At the same time the saturation characteristics are pretty fixed regardless of steel blend, at around  Tesla RMS, so one can work out the lowest dV/dt for which the steel appears to increase the inductance, even if the inductance for conditions less than this is hard to deduce.

At 50Hz you need several feet of conduit to become noticeable, and that is assuming conduit made of transformer steel (such as the stuff described around page 30 in that link),

regards Mike

PS example  BS 4568 conduit dimensions are 20 or 25mm OD, and 1,6mm wall thickness.

(magnetic path length of pi * 18.4mm for the 20mm, and pi* 23.4mm for the 25mm, magnetic area 1.6mm* length)

Let as assume the 2 extremes - a high mu steel of say 20,000

a 1m length of 20mm conduit would add  20000* 4π × 10-7 H/m* 0.016m^2/ (0.01844*pi) =  0.006956522 H

(6.9milli Henries.)

and a low mu case of 500.

then 0.000173913H or about 170 microHenries.

Then we should consider the saturation of that core - in the high mu case about 20A/m - and given the magnetic path is only about 50mm long that saturation occurs at around one amp in the wire.  For a conduit of  low mu steel, we an expect more like 40A.

For long runs of conduit the inductance becomes worth considering for its effect on low current Zs readings - after all at 50Hz  1millihentry is 1/3 of an ohm (more or less).

All likely fault currents will not see the full inductance or anything like - as shown above, somewhere between 1A and 40A the steel will saturate instead, and the extra inductance 'switches off'

For unknown steels it is fairly safe to assume the loss impedance is about 10% of the inductive impedance at most, and usually quite a bit less. - so you can imagine the effect of threading conduit over the wire not just to add inductance, but also to transformer couple into a loss resistor.

Mike.

IronFreely said:

is that a close enough for rock and roll just of things here?

Works for me in this case.

This reminded me of a IET discussion thread I started a while ago called

A Friday Debate

engx.theiet.org/.../a-friday-debate