[note - I wrote this after the 12:00 posting, before seeing later ones: it's about the general case of PNB with 4 conductors from a transformer, not about what things might be happening with the unknown connections in the OP's system]


Surprising and rather interesting.  I see John's point about BS7430 etc. From a following-standards perspective it indeed seems that "PNB implies TNCS" is the dominant view.  Below I note that standards seem to see a PEN in PNB too.


That said, I agreed with most of the reasons that have been given here for the above being a strange choice of definition.  In particular, people associate TNCS with a shock danger in the event of a single conductor fault without need of any other fault: that wouldn't happen for any broken conductor in a PNB system (excepting leakage currents). The argument that topology rather than length should be decisive (the cable from transformer to earthing point just "extends" the transformer terminals) is also strong.


Regarding whether there's a 'PEN' in PNB...  The BS7671 definition of PEN is simply about sharing protective and neutral conductor functions; and the definition of protective conductor functions includes connecting an earth electrode to a source neutral point. If the neutral point is the point where the three phases' neutral-ends meet (in or on the transformer), then the conductor in question in PNB is a PEN by this definition. Here, again, this would not have been my choice when thinking about the effects of a break in this conductor with no other fault in the system. Even a TT system could by this definition have a 'PEN conductor' if the neutral from the source runs as a wire to the first earth electrode and then on to the customers, as is often found in overhead supplies with separate HV/LV earthing where the LV earth is at the "next pole". 


Diagrams about multiple sources (rather than specifically PNB) in BS7671 2008 Figs.9A/9B have a note (b) that "the function of this conductor is similar to a PEN". I don't have a more up to date BS7671, but I see the Swedish 2018 equivalent keeps that diagram and much the same note, so it's not just a locally added choice of definition. If the standards believe their own definitions that I mentioned above, they should probably drop "similar to" and say "is a PEN". This diagram of multiple sources appears the most similar diagram to PNB, since the simpler diagram showing a single-source TNCS includes optional earth electrodes at multiple points without distinguishing the case where there are no electrodes on the source side of the PNB/N/PE/electrode point. The interest of the standards appears to be about whether the N and PE emerge separately at a place called a source, or a distribution supply cable, or whatever, rather than by where this point is relative to the earth electrodes.


Although it still sounds strange to me, I do see some mitigating reasons for the views of the standards now that I've been pushed to think about it:


Notation.

    If, as typically explained, we treat the 'T' of TNCS as meaning that "the source has a part [neutral] connected to earth" and the 'N' as "protective earthing of installations connects to the source neutral", then it's true that in PNB there's a conductor 'C'ombining both of these functions, besides other places where a 'S'eparate conductor has just the PE function. That's taking the fussy view that the source neutral is the point where the neutral ends of the separate source-phases meet. I like the "extending the terminals" argument in its way, but can't deny that a cable with its associated connections, extending tens or hundreds of metres from the actual neutral "point", introduces further opportunities for a break compared to a simple chunk of copper with bolts.  I don't think anyone has stated what (exact) basic definition of the TN* naming system we're using, or what we'd regard as the authority for this.


Safety.

   In the event of a fault from a higher-voltage system (interwinding fault in the transformer, or faults between different systems' conductors) the PNB neutral connection might be essential to reliable fault disconnection of the higher voltage supply, besides to limiting overvoltage on the loads during the fault. This is a bit of a cheat example as it's not the protective function we'd normally think of, and wouldn't arise if the source were a generator away from other electrical systems. However, it does give this piece of conductor a special significance beyond just normal operation.

   Without significant load, an earth fault (L-PE) in a PNB system with broken transfomer-neutral connection would not cause dangerous voltage on exposed parts in the installation, although it would cause harmful voltage on some loads. However, with plenty of load, there might be enough current through a L-PE fault to cause potentially dangerous touch voltage at the end of a circuit (rated much less than the installation load), while not giving good disconnection time. A little straw-clutching, but it's not out of the question, and indicates the conductor in question (PNB N-conductor) to have a role in earth-fault protection. Whether it's preferable to have a classic TNS system (earth at source) with broken main PE path is another matter.


Conclusion: "PNB -> TNCS", and "TN*-system source neutral connection = PEN", seem 'pandemic' in standards, and rather confusing regarding our usual first-thoughts about the implication of 'TNCS' and 'PEN', but not as strange as I thought yesterday. And I'll stop this legalistic study now, having been appreciative of the discussions here and of this little surprise. It almost certainly will never matter to me, but it's good to be warned yet again that much that seemed clear is not.

[note - I wrote this after the 12:00 posting, before seeing later ones: it's about the general case of PNB with 4 conductors from a transformer, not about what things might be happening with the unknown connections in the OP's system]


Surprising and rather interesting.  I see John's point about BS7430 etc. From a following-standards perspective it indeed seems that "PNB implies TNCS" is the dominant view.  Below I note that standards seem to see a PEN in PNB too.


That said, I agreed with most of the reasons that have been given here for the above being a strange choice of definition.  In particular, people associate TNCS with a shock danger in the event of a single conductor fault without need of any other fault: that wouldn't happen for any broken conductor in a PNB system (excepting leakage currents). The argument that topology rather than length should be decisive (the cable from transformer to earthing point just "extends" the transformer terminals) is also strong.


Regarding whether there's a 'PEN' in PNB...  The BS7671 definition of PEN is simply about sharing protective and neutral conductor functions; and the definition of protective conductor functions includes connecting an earth electrode to a source neutral point. If the neutral point is the point where the three phases' neutral-ends meet (in or on the transformer), then the conductor in question in PNB is a PEN by this definition. Here, again, this would not have been my choice when thinking about the effects of a break in this conductor with no other fault in the system. Even a TT system could by this definition have a 'PEN conductor' if the neutral from the source runs as a wire to the first earth electrode and then on to the customers, as is often found in overhead supplies with separate HV/LV earthing where the LV earth is at the "next pole". 


Diagrams about multiple sources (rather than specifically PNB) in BS7671 2008 Figs.9A/9B have a note (b) that "the function of this conductor is similar to a PEN". I don't have a more up to date BS7671, but I see the Swedish 2018 equivalent keeps that diagram and much the same note, so it's not just a locally added choice of definition. If the standards believe their own definitions that I mentioned above, they should probably drop "similar to" and say "is a PEN". This diagram of multiple sources appears the most similar diagram to PNB, since the simpler diagram showing a single-source TNCS includes optional earth electrodes at multiple points without distinguishing the case where there are no electrodes on the source side of the PNB/N/PE/electrode point. The interest of the standards appears to be about whether the N and PE emerge separately at a place called a source, or a distribution supply cable, or whatever, rather than by where this point is relative to the earth electrodes.


Although it still sounds strange to me, I do see some mitigating reasons for the views of the standards now that I've been pushed to think about it:


Notation.

    If, as typically explained, we treat the 'T' of TNCS as meaning that "the source has a part [neutral] connected to earth" and the 'N' as "protective earthing of installations connects to the source neutral", then it's true that in PNB there's a conductor 'C'ombining both of these functions, besides other places where a 'S'eparate conductor has just the PE function. That's taking the fussy view that the source neutral is the point where the neutral ends of the separate source-phases meet. I like the "extending the terminals" argument in its way, but can't deny that a cable with its associated connections, extending tens or hundreds of metres from the actual neutral "point", introduces further opportunities for a break compared to a simple chunk of copper with bolts.  I don't think anyone has stated what (exact) basic definition of the TN* naming system we're using, or what we'd regard as the authority for this.


Safety.

   In the event of a fault from a higher-voltage system (interwinding fault in the transformer, or faults between different systems' conductors) the PNB neutral connection might be essential to reliable fault disconnection of the higher voltage supply, besides to limiting overvoltage on the loads during the fault. This is a bit of a cheat example as it's not the protective function we'd normally think of, and wouldn't arise if the source were a generator away from other electrical systems. However, it does give this piece of conductor a special significance beyond just normal operation.

   Without significant load, an earth fault (L-PE) in a PNB system with broken transfomer-neutral connection would not cause dangerous voltage on exposed parts in the installation, although it would cause harmful voltage on some loads. However, with plenty of load, there might be enough current through a L-PE fault to cause potentially dangerous touch voltage at the end of a circuit (rated much less than the installation load), while not giving good disconnection time. A little straw-clutching, but it's not out of the question, and indicates the conductor in question (PNB N-conductor) to have a role in earth-fault protection. Whether it's preferable to have a classic TNS system (earth at source) with broken main PE path is another matter.


Conclusion: "PNB -> TNCS", and "TN*-system source neutral connection = PEN", seem 'pandemic' in standards, and rather confusing regarding our usual first-thoughts about the implication of 'TNCS' and 'PEN', but not as strange as I thought yesterday. And I'll stop this legalistic study now, having been appreciative of the discussions here and of this little surprise. It almost certainly will never matter to me, but it's good to be warned yet again that much that seemed clear is not.