Log in to the online community

Want to post a reply? You'll need to log in
Two supplies
lyledunn
371 Posts
Question
Two DNO 11Kv/400/230v transformers on different HV lines originally supplied two separate buildings on an industrial site. The two separate buildings are now one large steel framed one. So we have two DNO intakes and no clear demarcation within the installation. I remember reading somewhere about issues arising in such situations but cannot find the article. I am keen to establish if any issues are serious enough to seek alternatives. One of the intakes has a Biomass plant feed-in with g59 relay. Both intakes are circa 1MVA. 
11 Replies
On the odd occasion where I've come across this before, usually the DNO baulk at the idea of two supplies into one building (demise) and cite earthing as their reasoning.  As long as both the star points, one from each transformer, are solidly bonded together so there is one Earth potential, I don't really see the issue.  Have you spoken to the DNO?

As you have multiple supplies by the sound of it via the biomass plant, I guess there is already labelling regards safe isolation etc but that might also need revisiting based on having another TX supply in the single building.
mapj1
2351 Posts
There are almost two different things to consider. Firstly the LV interleaving, which may make isolation in an emergency tricky, and where earth fault currents flow, if there is a fault that returns via the wrong CPC.-  but that is true in any building with 2 LV supplies. (and in some cases one supply ends up having to be   TT and the other TN if they are neither separated nor tightly tied together such as second supplies for certain type of fire system)

Then the HV side - are these underground lines bringing in an HV earth on the armour with them and if not how are the HV earths arranged. Presumably they are supplied by the same DNO, but are they private transformers, or   if they are DNO property do they have an opinion as to how they are wired and how they should be interconnected or not .  I presume they are not side by side on a common ground plane.
OMS
693 Posts
Not usually a problem, if the HV (MV) side is coordinated - ie are they derived from the same source, or if not is each source synchronised. Then you just need the transformers to be of the same vector group.

You then have a common platform with common earthing and bonding.

As a fairly extreme example, a client of ours on a recent project has 2 x 60MVA supplies into one building from a pair of 33kV feeders (The 33/11kV transformers are within the curtilage but outside the building). He has 40 x 11/.4kV transformers and a potential for 60 x 1.5MVA Generators all connected at the same time - there are a number of earthing transformers in place as well as the generators act at 11kV

Speak with the DNO, but as it's likely that both supplies are from the same HV Feeder, than all they would want is a big earth cable between each intake (which could be the steel frame of the shed)

Regards

OMS
Hi OMS.

I had presumed, perhaps naively, that the supplies were not actually going to be paralleled but the OP does not state, so I hadn't considered vector group of the transformers.   

With that said, if the TXs were by default sharing a return path via steel frame or other earthing, would the different vector groups of differing TXs cause any zero sequence currents to be at differing angles?  Or have I just made that up?

Thanks,
mapj1:
 TT and the other TN if they are neither separated nor tightly tied together such as second supplies for certain type of fire system)

Then the HV side - are these underground lines bringing in an HV earth on the armour with them and if not how are the HV earths arranged. Presumably they are supplied by the same DNO, but are they private transformers, or   if they are DNO property do they have an opinion as to how they are wired and how they should be interconnected or not .  I presume they are not side by side on a common ground plane.

Hi.

Interesting that you mention TN + a TT supply, this has been suggested to me by IDNOs before, however, without any technical justification.  Is this approach due to the idea that the two earths (star points) will become disconnected somehow, so it's better to design it that way in the first place?  Just TT means an RCD and that is rather counter to the purpose of many secondary supplies.

As for how the HV is earthed, presuming its a cold site, I don't see the issue as long as the HV lines do not extend further into the building than the transformer room.  What am I missing?

Thanks

mapj1
2351 Posts
Not wishing to second guess the IDNOs, but a building with two TNC-S LV supplies is one of those things that needs to be done carefully 🤔
Because in effect you have two NE bonds, and two lots of bonding that put the neutral onto the structural steel, plumbing etc, when those neutrals join up round the back via a common low impedance street network ground / DNO neutral / HV ground there is potential to have a constant load dependent voltage gradient along the building. This is similar to the metal water main serving as a second PEN in a street of houses, but with less impedance to limit the current. Even though the voltages may be fractions of a volt the currents may be tens or in an unlucky case,  hundreds of amps, with all the associated magnetic fields and this is 'a complication' , and is is either avoided by tying both together very solidly - easy if side by side, less so at opposite ends of a barn, or more easily using the electrode resistance in a TT system is to reduce the potential for large circulating currents. 
In the end the folk who know the network details should be the DNO, but things like building change of use of a building that was separate modules into one big one, or the addition of factory piping running between previously isolated buildings can introduce problems where there originally were none.
Adha Azmir
10 Posts
For better solution view, provide the schematic due to not know the 1st design is. Is it still maintain the old ones or this is the new design. 
gkenyon
1167 Posts
Dutch of the Elm:
mapj1:
 TT and the other TN if they are neither separated nor tightly tied together such as second supplies for certain type of fire system)

Then the HV side - are these underground lines bringing in an HV earth on the armour with them and if not how are the HV earths arranged. Presumably they are supplied by the same DNO, but are they private transformers, or   if they are DNO property do they have an opinion as to how they are wired and how they should be interconnected or not .  I presume they are not side by side on a common ground plane.

Hi.

Interesting that you mention TN + a TT supply, this has been suggested to me by IDNOs before, however, without any technical justification.  Is this approach due to the idea that the two earths (star points) will become disconnected somehow, so it's better to design it that way in the first place?  Just TT means an RCD and that is rather counter to the purpose of many secondary supplies.
 

Guessing in this case the supplies are at LV. The issue for DNO/IDNO is circulating currents, if they make two N-E bonds (as mapj1 pointed out). Also from their perspective, of course they need to make it so one supply cable can be worked on if the other is live.


From the installation perspective, don't forget, if you have two supplies serving the same location(s), you need to be careful to comply with Regulation 411.3.1.1 - especially second para:

Simultaneously accessible exposed-conductive-parts shall be connected to the same earthing system individually, in groups or collectively.

So if you have two supplies, you must either segregate them physically at all points (even if exposed-conductive-parts are shared), or bond them together ... but see Reg 542.1.3.3.
 

mapj1:
Not wishing to second guess the IDNOs, but a building with two TNC-S LV supplies is one of those things that needs to be done carefully 🤔
Because in effect you have two NE bonds, and two lots of bonding that put the neutral onto the structural steel, plumbing etc, when those neutrals join up round the back via a common low impedance street network ground / DNO neutral / HV ground there is potential to have a constant load dependent voltage gradient along the building. This is similar to the metal water main serving as a second PEN in a street of houses, but with less impedance to limit the current. Even though the voltages may be fractions of a volt the currents may be tens or in an unlucky case,  hundreds of amps, with all the associated magnetic fields and this is 'a complication' , and is is either avoided by tying both together very solidly - easy if side by side, less so at opposite ends of a barn, or more easily using the electrode resistance in a TT system is to reduce the potential for large circulating currents. 
In the end the folk who know the network details should be the DNO, but things like building change of use of a building that was separate modules into one big one, or the addition of factory piping running between previously isolated buildings can introduce problems where there originally were none.

Why is bonding them together hard if the supplies are not adjacent (but still local to the same building)?  Is this not just the case of a large section bonding conductor between them, and making sure they stay bonded?  Maintaining bonding throughout is already an important safety aspect, so I don't see why we should assume this would not be done. 

lyledunn
371 Posts
Thank you gentlemen. The Supply arrangement are existing. Both transformers are dedicated and have no other consumers. The HV arrangements are unknown. They are DNO transformers and supplies are taken at LV and are TNS. I am not really up on G59 but I thought that the relay disconnects the AD plant input at the panel in the plant. If the two supplies were inadvertently interconnected say by a mistake in a simple ring final circuit, it would appear that the busbars intended to be disconnected would be live?
gkenyon
1167 Posts
Dutch of the Elm:
mapj1:
Not wishing to second guess the IDNOs, but a building with two TNC-S LV supplies is one of those things that needs to be done carefully 🤔
Because in effect you have two NE bonds, and two lots of bonding that put the neutral onto the structural steel, plumbing etc, when those neutrals join up round the back via a common low impedance street network ground / DNO neutral / HV ground there is potential to have a constant load dependent voltage gradient along the building. This is similar to the metal water main serving as a second PEN in a street of houses, but with less impedance to limit the current. Even though the voltages may be fractions of a volt the currents may be tens or in an unlucky case,  hundreds of amps, with all the associated magnetic fields and this is 'a complication' , and is is either avoided by tying both together very solidly - easy if side by side, less so at opposite ends of a barn, or more easily using the electrode resistance in a TT system is to reduce the potential for large circulating currents. 
In the end the folk who know the network details should be the DNO, but things like building change of use of a building that was separate modules into one big one, or the addition of factory piping running between previously isolated buildings can introduce problems where there originally were none.

Why is bonding them together hard if the supplies are not adjacent (but still local to the same building)?  Is this not just the case of a large section bonding conductor between them, and making sure they stay bonded?  Maintaining bonding throughout is already an important safety aspect, so I don't see why we should assume this would not be done. 

What cross-sectional area do you put on that "large section bonding conductor"? See Reg 542.1.3.3 . Don't forget, though, the current the conductor has to carry may not simply be fault current (whether or not originating from the LV or HV network), but may have to include Neutral currents - and perhaps not just Neutral currents from the installation(s) in question.

Share:

Log in

Want to post a reply? You'll need to log in