Joined: 01 October 2007
I am currently working as site engineer on a large windfarm in central asia. I am trying to work out why the grid is insisting that the HV neutral of the main windfarm export transformers be run disconnected from the HV neutral earth they originally insisted on.
The system is connected to the grid via two YNd11 110 / 35 kV 50MVA transformers giving N+1 redundancy. This connects to a double circuit 110kV OHL circa 35 km long to a DNO substation connected to 220kV via YNYn0 transformers.
The windfarm transformers were originally specified to be Dyn11, however, the local grid were insistent that they wanted an earth at both ends of the connecting transmission line. Their normal practice is to use YNyn0 transformers throughout the grid, and run everything at 12 o'clock.
I recognise that normal UK practice is earth at both ends of a line at 132kV and above, but for 66kV earth at source is preferred and this would have been the case with this system (110 kV could be treated like a UK 66 or 132kV network if falls between the two in my opinion).
My colleague who specified the original transformer wished to reduce the costs by not requiring a tertiary delta to filter triplen harmonics (therefore removing need for a 5 limb transformer saving 25% costs), and ensuring that the 35 kV windfarm system had an earth at the star point of the secondary. We are aware that a star connection on the HV would have been cheaper as graded insulation would have been possible, however, due to the 35 kV network setup and a requirement for a 35 kV earth it was decided to not do this, as this would have required additional earthing transformers.
The grid was insistent that they would require a solid HV earth and after much deliberation, the system was redesigned to provide YNd11 main transformers with a 35 kV earth transformer.
I am currently on-site as client's site engineer. The grid's switching operatives are insisting that the transformer earth diconnectors are closed in when the transformer is energised, but then removed for normal operation.
Can anyone think of a reason why the grid would wish to run with the earths disconnected, having insisted on them to start with?
From a protection point of view to my mind, having an HV Y winding with no earth is not different from having a system with an HV delta. The only way to reliably protect this system would be with NVD protection on the HV side to detect broken line and faults where a line is down to earth at the client side. I have simulated this in Matlab and can't find any reason why an unearthed Y is any better than a D winding.
I would be interested to know everyone's opinion.
Joined: 09 September 2002
I was hoping to read a knowledgeable response to your post, as I don't have a good answer for you. In the absence of such, I offer the following comments:
1 - You don't specify the Country - in my experience, in countries other than the UK (in particular Portugal & Turkey), DNOs have policies (frequently enshrined in law) that don't seem to have good engineering reason behind them. Sometimes they have fixed views on vector groupings that would only be relevant if they were incorporating your 35kV as an alternate feed to their network.
2 - It would seem that there's either been a mistake by one or other (or both) of the DNO engineers.
3 - I can't see the virtue in earthing your 110kV side during energisation only. However, if the switching device is a disconnector, then I'd be looking to block its operation if any significant current is flowing that may cause overstressing or failure.
4 - If you went for graded insulation on the 110kV side, it may be worth considering installing a surge arrestor N - E to protect this under fault conditions.
5 - I don't see that NVD on the 110kV side will protect your transformer, though it would provide E/F protection on the DNOs side ie, prevent your installation from continuing to feed their E/F from a broken conductor.
6 - For a transformer of this size, in addition to Buchholz ryc, I'd expect its protection to include Bias Differential, HV OC + Balanced E/F (if N-E intact during energisation, you could incorporate HV REF for the cost of an additional CT). On the LV side, OC, REF, Balanced E/F and Standby E/F.
2 - A bit late, but I would have gone for a Yd configuration to balance out HV voltage imbalance being transferred to the LV side. I know this requires 35kV earthing transformers, but these can be fitted with a secondary winding that can provide Aux supplies to your S/S. As you say, this vector group would also prevent triplen harmonics crossing the transformer in either direction.
I hope these comments might be helpful, but I will be interested to read contributions from others.
Joined: 06 February 2013
More shots in the dark?
Is it in a desert?
What protection is being used at the feeder end?
They might be worried about an o/c conductor near your end of the line that faults to earth but only on the non feed side.
They might think you have a breaker which would then trip on e/f beating their protection, isolating the fault and leaving them unaware of the fault (at least until you tell them)......
Why not ask them? (and let us know)