Earthing of Neutral in Substation or MCB at LV Panel ? or both ?

The answer is that both are done, but in different circumstances. Also I'm far from expert on this, so wait for other to chip in. There are some questions that will help clarify the situation.


1) Is this in the UK? - if not the answer may change - here there are special rules about multiply earthed neutrals and other countries may do different things with HV side earths and fault detection.


2) How is the HV earthing to be sorted ?

If the site is "cold" (DNO may be needed to advise on characteristics of the HV side to determine this) then the HV earth -i.e. the transformer core and tank, and the LV earth i.e. the star point of the transformer and your PEN, may share earthing arrangements, and indeed if it is possible, then  it  is preferable if they do.

If the site is "hot" - that is to say a credible HV fault current could provide enough current to raise the HV earth to a dangerous voltage, then the HV and LV earths must be separated systems, and must not be co-located,

If the LV earth needs to be remote from the HV one where is it best physically?


How far apart are the TX and the building with the breakers an so on?


Mike.


Mike.

Hi Mapj ! 


1) This is not in UK but I have found this forum very helpfull and people with lot of knowleadge that might help me !

2) In the substation there is the earthing grid below the substation. I dont know what you mean by hot or cold.. 


The substation is 80 meters away from the main circuit breaker and might me like 180 meter from the las circuit breaker.


The contractor suggested me to ground the neutral only at the main circuit breaker, and not at both sides.. 


Also to tie the earth grid of the substation to the LV earth where the neutral is going to be grounded.. (80 meters away one from the other)


They will be tied with a 240mm2 Al cable.


Whats your opinion about this ?

I dont know what you mean by hot or cold..

It's an old UK rule of thumb for deciding whether the HV earth and LV earth are combined or kept separate. Generally if there's a metallic path back to the HV source for earth (or some other means of keeping the HV earth below 1Ω) (cold site) then the earths are normally combined. Otherwise (hot site) the HV and LV earths were kept strictly separate so HV faults weren't imposed on the LV system.


I gather that these day the process is a little more sophisticated and a proper calculation for the rise (and duration) of earth potential in the case of an HV fault is done.


   - Andy.

The contractor suggested me to ground the neutral only at the main circuit breaker, and not at both sides

I think it's usual to earth the N at just one point - as that tends to produce better EMI results as there's less current flowing unbalances around random metalwork or soil. In the UK it also avoid legal complications of a "consumer" having a combined neutral and earth conductor in their installation.


Whether N is earthed at the transformer or switchboard end tends to vary depending on the circumstances. With one one consumer DNOs tend to favour earthing the N at the consumer's end (PNB) as that reduces the potential difference between N and PE for the consumer. Clearly where the HV and LV earths have to be kept separate it's likely to be convenient to earth the LV N away from the transformer, if they're combined it might well be easier to earth everything together at the transformer.


  - Andy.

Sorry - as above the 'hot' / 'cold' thing is about HV side faults.


In the UK the HV neutral is not distributed, there are just the HV phases, and underground HV will have an earthed armour, but not at zero impedance to the HV network, there may be current limiting by a neutral earthing resistor of typically about 1 ohm.

So any HV  fault to earth has defined current of (say 6000A or less with that resistor of 1 ohm and 11kV phase to phase 6k6 to ground), and can be detected and interrupted promptly by what is in effect  giant sized RCD on the HV supply made from discrete current transformers and dedicated breaker firing circuits.


On an overhead supply the fault current is generally limited further by the path through terra-frma,


Once you know the HV fault current, you can look at the substation earthing arrangement and decide how much voltage rise an HV fault from say one phase to the transformer core would cause, and how long it would last for.

This is where in the UK the DNO netork planners are are needed to advise.

Then you can decide if you would dare to put that over-voltage onto the LV network. If the answer is low enough that it is safe, then the site is 'cold' and HV and LV earths can be combined.

If not the site is 'hot' and HV earthing (the transformer core and case are considered at single fault risk of being live to  phase voltage, while the secondary windings would require multiple faults to reack HV voltage)


The internals of a UK transformer keep the HV and LV sides apart and off earth.

The photo below shows the details of typical connections - it's not my photo, but the flat 'ribbons' are the outgoing LV, and the thinner wires with the more insulated feed-through bushes are the HV side.


regards Mike


Edited to actually attach the photo this time - my apologies !!

Normally this would be oil filled, and there would not be a human arm left in the tank when it is eventually energised  (!) - but it gives an idea of the size of the insulators and the arrangements..

The main thing is that the 3 HV bushes coming in on the left, and the 4 (3 phases and an N) low voltage high current taking power out  the right . An earth goes direct to a bolt or stud on the outside I think this is a 500KVA (half megawatt ) unit, but the TX itself is out of view, to the bottom of the picture, and the wattage is not that important here - the thing  I want to show  is the neutral is as reliably isolated from the transformer and tank as the 3 phases, so HV and LV earths are very separate at this point.