Minimum would be 0.0V. Maximum could be almost anything (up to Uo). The classic problem with TT systems is that during an earth fault, all the fault current returns via the source electrode, multiplied by the soil resistance, can give a very substantial potential difference between true Earth and the transformer's star point (i.e. supply N) -  the exact figure depending on the resistance of the consumer's electrode (and any parallel paths - e.g. bonding to extraneous-conductive-parts) as much as the source electrode itself. If the fault isn't cleared promptly (e.g. due to a faulty RCD or just because it's on the distribution network) that N can reman at a hazardous voltage for a long time. Hence all isolation devices need to open N as well as L on TT systems.

  - Andy.

Actually, in a  building continuously indicate  neutral to earth voltage  as 10 to 11 V. Main cause for this is harmonics of VFDs.  Hence , I want to know is there recommended maximum voltage levels in-between neutral and earth of TT system (with relevant standard reference). (In above mistakenly I wrote a minimum now I'll correct same.) 

There are no recommendations of N-PE voltage from safety standards.

However, there are recommendations for EMC (although not from standards) - which have, in some cases, been as low as 5 to 10 V.

There is a solution to this ... although not without cost ... which is to use transformers to permit the filters to perform as they would in a TN system. This is recommended in the standards for VSDs, and also in BS EN 60204-1. The transformer secondary neutral is earthed, thus forming TN-S circuits in the TT installation.

On a single phase system, 11v is probably on the borderline of being acceptable. In contrast, by the time it reaches 20-50 you almost certainly have an unrealised  fault somewhere. The thinking is as follows - the voltage drop in the LN loop may be right at the limit, as high as ~ 10% of supply (20 odd V) Half of this is lost on the live on the way out, and the other half in the neutral wire on the way back. TN-S would be the same, for the same reason. However this is neutral to true earth voltage - there may be an additional offset between true earth (an electrode far away from the near field of your local one and carrying negligible current) and your TT installation electrodes - if there are significant currents running into the electrodes then the metal work of the whole installation may also be a few volts off terra-firma earth voltage. It may be worth installing a test electrode, or even a garden fork or un-insulated screwdriver in the ground a few m from the building and seeing how far off the building earth voltage  that is.

Be aware in a TT system can go a bit wild if your user electrodes are lower resistance than the ones at the substation.  -  if this happens, when the fault to earth comes on, the voltage is divided between the series connection of your electrodes, and the substations transformers electrode(s).
Normally we assume the substation neutral voltage rises a few volts, and most of the 230V  or whatever, is dropped in the earth around your your local one, but it can be the other way about - so that the phase you have shorted to ground is nearer true earth if your earth is good enough, then the the whole transformer, neutral and everyone else's supply  bounces up to near 230V for the neutral and upto about 400V for the other phases. So while you have the test electrode in the lawn or flower beds, also check the phase voltage/voltages. you may not have a fault, but your neighbours just might.

Mike.

in a  building continuously indicate  neutral to earth voltage  as 10 to 11 V. Main cause for this is harmonics of VFDs.

There could be a number of things happening here -  protective conductor currents from the installation will raise the potential on the consumer's electrode to some extent (but I'd be surprised if it reached tens of volts under normal circumstances). In the simple case BS 7671 puts a limit of 50V on that before protective devices (RCDs) are expected to trip.  Reducing the resistance of the consumer's electrode can help.

There's also voltage drop along the supply N - (which will appear as an increase of N voltage at the load end) - in the UK with a supply under the ESQCR the maximum drop would be 16% or 230V (or from 253V at the source to 216.2 at the intake or a drop of 36.8V - around half of which would typically be along the supply N - so max 18 or 19V. Plus any voltage drop within the installation - so maybe another half of 5% - so even in an entirely compliant situation you could in theory be looking at N at a load being more than 24V adrift from Earth.

Then the supply reference (star point) can be held away from true Earth by Earth current returning via the source electrode (which is also influenced but protective conductor currents, if normally to a small amount).

Typically things aren't quite as bad as that though - if the source is 3-phase then N currents and protective-conductor-currents may well cancel out to some degree (even if your installation only uses one of the phases).

   - Andy.

AJJewsbury said:

In the simple case BS 7671 puts a limit of 50V on that before protective devices (RCDs) are expected to trip. 

Just to be clear, there is no limit in BS 7671 on touch-voltage to the general mass of Earth in this regard ... only in terms of selection of protective devices and combined resistance of protective conductor and earth electrode alongside this. There is no limit to touch-voltage other than that provided for by supplementary protective bonding, which will be local touch-voltage, and limited to either 50 V AC/120 V DC (or 25 V AC/60 V DC in medical locations).

Just to be clear, there is no limit in BS 7671 on touch-voltage to the general mass of Earth in this regard ... only in terms of selection of protective devices and combined resistance of protective conductor and earth electrode alongside this.

Which are explicitly co-ordinated so that a touch voltage of >50V cannot persist - reg. 411.5.3 (ii).

The touch voltage can of course exceed 50V for very short durations, but that's the same with supplementary bonding where the fault current exceeds Ia.

   - Andy.

AJJewsbury said:

Which are explicitly co-ordinated so that a touch voltage of >50V cannot persist - reg. 411.5.3 (ii).

Not disputing that, although only true in some circumstances, such as RCDs in TT systems ... it's unlikely that a high touch-voltage will persist in other cases for faults in the installation, ignoring PE faults and leakage currents.

The reason for pointing this out, is that there's a common misunderstanding that BS 7671 actually serves to limit touch-voltage values (although as you correctly point out, for the most part, it's aims to limit the time those touch-voltages may persist).