That was my point - that in the 'usual' situation, of splitting the PEN into N and E, the 'E' part of it does not go through the meter.
I think we're violently in agreement here. But I think you missed my point that if the supply (or rather, what's downstream) is TT, then any earth fault current wouldn't be going through the neutral connection - regardless of where any PEN got split into N & E. But I agree, it doesn't look a very good arrangement - at least in the absence of more information about the installation.
In any event, there may be a 'simpler', regulatory, reason why a PEN is not 'allowed' to go through a meter. I'm not totally sure (even having looked at the regs) exactly where "a consumer's installation" is deemed to start, but I would presume that it cannot be any further downstream than the output terminals of the meter. Hence, if what came out of the meter were a PEN, that would be in violation of both BS7671 and ESQCR.
Indeed, which is why it's probably rather fruitless to debate further in the absence of more information on what's downstream - I would imagine "not a standard supply as we are used to it".
Having said all that, as you go on to say ....
If the DNO have not provided a TN-C-S earth (i.e. haven't connected the installation's earthing conductor to the PEN upstream of the meter - or, at least, provided provision for that to be done), then what will go through the meter is 'just the neutral' - but it could become a ('not allowed') PEN if someone (not the DNO) subsequently decided to derive the installation's earthing from the post-meter neutral.
Indeed.
Indeed - it is, at the very least, potentially confusing (and would probably have been less so had the neutrals from all three of the outgoing circuits from the RH cutout been derived from the LH cutout).
However, even if the cutouts do have fuses in them, the neutrals of the two cutouts will be directly connected by the linking cable - so I suppose that (cable CSA allowing), electrically speaking there would be nothing wrong with what has seemingly been done.
Linking cable ? Do you mean the point where the supply cable neutrals get connected to one common neutral in the distribution network ?
Apart from potential issues around neutral currents (which I suspect aren't going to be significant in this situation), what happens in one of the cables gets cut ? If it's the left hand cable, then 3 flats lose their supply, and another loses it's neutral. If it's the right hand cable, then 3 flats lose their supply - so no great deal.
But in the latter case, do the other 3 flats also lose their earth ? I see there's one earth cable coming from the right hand service head - if that's the TN-C-S earth for all the flats then cutting one cable removes the earth for all 6 of them
But if the TN-C-S earths are derived at the individual service heads in the flats, then cutting the left hand cable not only removes supply from 4 flats, but also leaves a situation where any loads in one of the flats (the one with the cross-connected neutral) will apply 240V to the earths of those 4 flats
While there is always scope for losing a neutral while phase(s) remain connected - this seems to be a case of having created a higher risk of that by splitting phase and neutral between two cables that may be following different paths. Hard to tell, but it looks like there may be a mix of old and new there - left hand looks like old (painted ?) cable and new service head (outline of old head in the paint), right hand looks like new cable and new head. Normally if the proverbial digger bucket goes through a cable, it stands a good chance of breaking all the conductors -
relatively safe for the people downstream. But if it breaks one of the two cables then it's created the risk described.
Again, a case of "more info needed". But IMO it would have been significantly safer had the fitter used two Henley blocks - one for each side.