So, as they are private transformers, are the secondaries wired as TN-S ? Or if they are not could at least one of them be (or PNB, which achieves the same end)?

The earths must be interlinked, and may well be already - it is important that there is no chance of anyone straddling two earth zones at different potentials, and in this case there is no way to keep them far enough apart not to be simultaneously within reach.

If you are not paralleling the secondary windings, then apart from making it really clear where various things are isolated, then  there is no sensible current path that will give a diverted neutral problem, unless you had that already when they were separated.

It will be necessary to keep the circuits from the two transformers apart, and a very good idea to avoid things fed from the two ends from sharing panels or switchrooms or very very clearly marked if they must.

I'm sure others will chip in with further observations.

Mike.

The TN-C-S or TN-S grounding system, which provides protection against electric shock, is recommended for grounding steel-framed buildings. In this system, the grounding conductor and the neutral conductor are separated, and the building ground is connected to the grounding conductor and the neutral conductor to the neutral conductor. Since both transformers share a common grounding system, both buildings are considered one ground as required by 542.1.3.3. This means that all protective devices, such as fuses and circuit breakers, must be selected for the maximum load of both buildings. If a TN-C-S grounding system is used in steel-framed buildings, circulating neutral currents may occur, which can cause EMC problems. In such a case, the use of a TN-S grounding system, which prevents circulating neutral currents, is recommended. It is also important to make sure that the earthing system has a low enough earthing resistance to provide effective protection against electric shocks. Regular ground resistance measurements are recommended to monitor the quality of the grounding system.

If the building(s) are steel framed and all interconnected, then it sounds like you have a single earthing system whether you like it or not, so probably makes sense to design it that way from the outset. As Mike says, even if the earthing system is common, the neutrals should be separate (unless paralleled) as long as you adopt a TN-S rather than TN-C distribution arrangements - which you're probably obliged to do anyway under the ESQCR if the whole LV side is owned by the "consumer".

I think 542.1.3.3 is aimed more at the situation where small c.p.c. circuits might create a bridge between separate buildings that could be on different earthing arrangements  - think for example an outside light between two buildings with a requirement for 2-way switching with one switch in each building. Normal rules may well require the c.p.c. to be interconnected to both buildings earthing systems - and so the c.p.c. might then be obliged to carry very substantial currents between the two earthing systems (under fault conditions, but also diverted N currents under normal conditions if PME is involved). Earthing conductors associated with a supply are likely to be easily big enough to cope with any fault currents (or if the two supplies are very different sizes, the smaller might possibly need a small upgrade in size if it may carry a large proportion of fault currents from the larger installation).

100m does feel to be a long length of conductor to interconnect things - others may have a better idea of what's normal for this kind of thing. Maybe look to having a single "MET" at the building end (even if that's two terminals and a shorter interconnecting conductor).

   - Andy.

Hi Mike, thanks some good points. 

The secondary from each transformer is TN-S. The earth mat at each transformer are linked, therefore my thinking is that they share the same potential. We will verify this on site via testing between. So effectively earth on the MV side and by default on the LV side are one and the same. 

Secondaries are not paralleled, each transformer feeds a separate switchboard at opposite ends of each building so there is physical separation and circuits from each switch board are suitably protected back to their source. 

Signage, warnings and labeling is covered and will be detailed in the O&M's. 

Each Separate Sw Bd should have the same earth potential, hence outgoing circuits common in that respect and protected individually as you'd expect for any normal installation. I am currently of the opinion to not link the earths on the LV side as that could lead to issues with earth leakage, and possible neutral currents from one to the other, appreciate your thoughts on this? 

Thanks Scott, I'd agree, and it is the case that both 'supplies' share a common ground/earth. The system is TN-S. 

We should have impedance readings from then the earth mats and connecting cables were installed. We shall also check on site resistance between the two systems on the LV side which should give us a better indication of any issues. 

Hi Andy, thank you for commenting. Similar to other responses the system is TN-S. Understand the reasoning in 542.1.3.3 for final circuits, I am comfortable in this case that the two buildings will share earthing characteristics and are suitably protected. However, just didn't fully understand the rationale to earth within one installation only. It would be almost impossible to insulate between the two earthing systems in this instance. 

Would you still suggest MET's were interconnected (LV side) given the transformer earthing mats are interlinked with bare buried conductor? 

Two TNS transformers is very good news, as the diverted neutral current /TNC/ PME  questions now vanish.

If you like it or not the two LV earths will interconnect in a single building, if only via plumbing and building steels, so there will be cases where if a fault occurs in a fitting in the middle of the building, when current flows from the live of TX1 and into the earthing of both TX2 and TX1 - this is to be expected. It is therefore essential to make sure that it can get back to TX1 from the TX2 earthing, and that the largest credible fault current does not cause a problem if/when this happens.

I know youd do not really want it to take the long path, but some fraction of the current will do so - it will split in the ratio of the resistance, well impedance really,  of the two routes. Final circuit CPCs you can separate, but the incidental paths of things screwed to the wall or plumbing cannot be avoided, and so the METs are much better solidly linked than half-heartedly so via something that might not handle hundreds of  amps quite so gracefully. (what is the LV PSCC actually ?)

This may mean adding additional deliberate bonding between the two 'MET's. or you may have a large enough cross-section of steel/copper already.

Mike

However, just didn't fully understand the rationale to earth within one installation only. It would be almost impossible to insulate between the two earthing systems in this instance. 

It's just about protecting individual conductors. In our 2-way lighting example, you could take the cable into a Class II switch in the 2nd building and just terminate the c.p.c. to thin air (along with suitable precautions to prevent the switch being replaced by a Class I accessory). It's an option often chosen for things other than mains too - e.g. screens on signalling, data  or audio cables (Audio types often like to earth one end only anyway even within an installation to reduce mains hum) - the alternative would be a by-pass protective conductor in parallel to the screen(s) to shunt the bulk of the current past the screens. (Or use optical cables or wireless to design out the problem!)

Would you still suggest MET's were interconnected (LV side) given the transformer earthing mats are interlinked with bare buried conductor? 

Depends how well you can segregate what's supplied b the two different supplies - if you effectively have one installation spanning the entire building then 411.3.1.2 would demand a single common MET. If on the other hand you could keep exposed-conductive-parts of each side out of reach of each other (even if you had bonded extraneous-conductive-parts in common) you could treat it as two separate installations within the same building (like a terrace of houses). There again the steel frame might do the job of the MET for you (if sufficiently well interconnected) - making any extra bits of copper all a bit academic.

   - Andy.

 Rich Mc, You haven't said for either transformer how the earthing of the neutral on the LV sides are arranged (only that they are bonded on the HV side).

The risk that Reg 542.1.3.3 is trying to alleviate is that of currents from the HV side being diverted through the installation.

Are the HV and LV earths interconnected at each transformer? If so, when you connect them back together in the building, you get HV fault currents as a possibility in your building, which is undesirable, and for that reason Reg 542.1.3.3 ... in this case, a DNO would usually provide an earthing terminal for one of the supplies, but not both (meaning you need to TT one or both - which is an option regardless, except for additional cost of protective devices, of course).

However, if the HV and LV earths are not interconnected, and the buildings/installation are kept a suitable distance from the transformer earthing system influence (e.g. > 10 m from the earth nests and other buried/accessible conductive parts that are affected by transfer potential, although this could be less depending on the soil resistivity profile and HV fault calculations) then you make a new LV earth nest for the LV supplies building, and then the systems DO share a common earthing arrangement and the two supplies can co-exist as TN-S in the same building. This would be what GN8 refers to as a private 'PNB earthing arrangement'.

Relevant experience with BS EN 50522 and BS EN 61936-1 are required to make the decision on this one.