Incomer Protection Requirements where LV cable is direct from transformer?

Question

where the LV cable from the transformer is directly terminated at the transformer LV termination box i.e. no circuit protection on the LV take off chamber and feeding into our LV ACB incomer downstream. what is the requirement for circuit protection specification at the LV ACB here? We have been told that the LV earths need to be separate from the HV earth, so we've considered the system to be a TT earth from the LV ACB downstream and will include ground fault settings as an integrated part of the CB specification. 
 However do we need to consider the cable between transformer and ACB as this would be unprotected and not subject to earth fault as we have no cpc between transformer and ACB panel? is this section of cable subject to any '' IT '' considerations? 
 quick sketch to try and make clear what i'm on about :) 
 thanks

mapj1 · Accepted Answer

The requirement to keep the earths apart usually relate to what is in effect a high PSSC on the HV supply side, and the impedance of the local soil. The problem is that a HV phase to transformer core fault for example, will raise the core, the HV earth electrodes and anything connected to it to a dangerous potential, in the time before the HV side ADS does its thing. If this voltage is higher than the LV wiring can safely withstand - and what voltage that is, is rather open to interpretation, but lets say it is typically 600volts, then it is preferable for the LV side earth electrodes to be sufficiently far outside the near field of the HV side earth electrodes to stay well below that - which in practice may require it to be tens of metres away. For small 'pole pig' transformers where the HV earth is always a bit doubtful, this generally means the LV earth is one pole along.. Similar thinking applies for floor pad transformers, but it is less obvious. 
 It is still TNs or TNC-s on the LV side though - even if the actual electrodes are not co-located with the transformer as they still do connect via the neutral to the transformer star point. 
 To combine the two earths is preferable, if, and only if, the electrode impedance to terra firma can be kept low enough to avoid this 'rise of earth potential' problem- which ends up typically requiring an ohm or less, (*) and that means a substantial 'nest' of buried metalwork and good conductivity ground - which is just impractical to achieve on many sites. regards Mike 
 * for a long time DNO advice/rules was simply that an ohm or less meant a cold site, and HV and LV earths could share one set of electrodes, and more than an ohm meant 'hot' and independent primary and secondary earths kept well apart. Now we have maps of soil conductivity and better models, for both fault current and seasonal variations, so the calculation process is more refined, but in most cases the outcome is the same !

AJJewsbury · Answer

To my mind, not TT (you need two separate electrodes, one for the source another for the consumer to make a TT system). Not IT either. 
 If the cables are armoured, then the armour forms the c.p.c. Draw out a full diagram - you should still see a complete "fault loop" for a fault between a live conductor and the armour - via the N-PE link (whether it's at the transformer or your panel) - it doesn't really matter that the current has to flow though some conductors that are "downstream" of the fault - as long as the fault current can flow.in a complete loop. The HV fuses (or CBs) should see the fault current (if scaled down by the winding ratio of the transformer) and if it persists long enough, disconnect. 
 - Andy.

Incomer Protection Requirements where LV cable is direct from transformer?

Community Rules & Guidelines