The cables are 'cold'. Using the 70° values will give inaccurate results.
Sure, in terms of this exercise, that's true. However, I was asking about what figures/calculations electrician's 'routinely use', in the real world, when one's interest is likely to be in the 'worse case' in-service VD or Zs - in which case, as I said, it would seem appropriate to use 'operating temperature' figures (even though, as you say, those temperatures are probably rarely attained in practice). In any event, as you go on to say ....
In my mind, that is why the maximum allowed Zs is reduced to allow for temperature difference rather than adjusting the measured figure.
I had forgotten that the regs themselves and the OSG take opposite approaches (which must confuse some people) - but they
do both end up considering the situation at 'operating temperature'. The regs quote 70°C resistance/VD figures, but 'uncorrected' 'maximum Zs' ones (relating to those 70°C resistivities), whilst the OSG assumes 10°C (not 20°C, I'd forgotten that) resistivity figures but then gives 'adjusted maximum Zs figures' to take into account that resistance relates to the lower temperature. In theory, those two approaches should end up with much the same answer (give or take the 10/20°C difference). Taking the present example, the difference in resistances at the two temperatures is about 1.20 (23.5/19.51) and the 'correction factor' used for the OSG Zs figures is 1.24 - so much the same. I'm sure that some must get confused by this difference between BS7671 and OSG, but, as I said, both end up (by different 'routes') effectively giving the maximum Zs at 'operating temperature'.
I don't know why the figures are lower than you expected. Does the relevant volt drop not equate to 19.51 x 1.2? That is if you were using the cpc.
It was only the difference between the 20°C (R1+R2) (i.e. 19.41 mΩ
and the corresponding 70°C figure (~23.5 mΩ
that felt larger 'than I expected' - but that was only because I was foolishly relying on 'intuition', rather than calculations or tables
Kind Regards, John