Bonding earthgrids together 3

[veritas]

I have two earthgrids A and B. A is a power station with an 11kV 30MW generator and a 11/22kV step-up trfr. There is a 22kV cable from the trfr to a swbd on grid B 40metres away. Grid B is fed either from the power station or utility via a 132/22kV transformer situated on Grid B. Earthgrid B has been designed for a 132kV fault from the utility.

1) What fault level should Grid A be designed to?

2) Should grids A and B be joined or left seperate.

Thanks in advance.

 

[Marmite]

Where earthing systems exist in proximity to each other, but are not combined, part of the EPR from the HV system will be impressed on the LV system. In terms of earth grid separation, 40m isn't much separation for the size of grids you are talking about. I would bond the grids together. The advantage of bonding them together is that it will reduce the overall earth grid resistance because the earth grid area has a dominant effect. The disadvantage is that elevated voltages are transferred further afield. In major substations comprising both zone and transmission substations on the same site, combined earthing systems are always used. If you don't bond them together you may have problems with transfer voltages from pipes, fences, underground power cables, low voltage neutrals, communications circuits, crash barriers etc.

Regards
Marmite

 

[veritas]

Marmite

Thanks for some very useful insights. What fault current would you use to size the power station earthgrid? The way I see it any EF on the power station will not lead to an EPR as the EF current flows in the actual grid. So what do you use?

Thanks.

 

[Marmite]

The earth grid conductor sizing depends on the maximum earth fault current (originating from either system) and the protection clearance time. Normally the backup protection clearance time is used. The EPR depends on the current flowing into the general mass of the Earth through the earth grid, governed by the earth grid resistance. If you have a local transformer within the earth grid, earth fault current will flow into the grid, heating up the earth grid conductors and flowing back to the neutral, and therefore it has a bearing on the earth grid conductor sizing.
However, in your case the 132kV fault will more than likely be worst case from an EPR perspective.

Regards
Marmite

 

[Marmite]

For clarity, I should say that when sizing earth grids the worst case fault current is typically used, which is likely to be the three phase symmetrical short circuit current.
Regards
Marmite

 

[veritas]

Marmite - surely it cannot be a three-phase fault as by definition the currents sum to zero so no current to ground?

It makes sense to me that the conductor sizing can only be due to a 132kV earthfault (btw the 22kV system is EF limited to 25A) BUT a 132kV fault cannot possibly give rise to an EPR since the 132kV winding of the trfr is delta connected. Though a 132kV EF close to grid B could have some EF current flowing to the grid an into the OHEW of the transmission line? But how to quantify?

Sizing of Grid A would be due to an EF on the 11kV side as the 22kV is limited to 25A.

 

[7anoter4]

I agree with Marmite. The short-circuit current which has to be considered it is the maximum short-circuit in the worst situation[it could be for 132kV system for phase-to ground-to phase-double ground fault].

 

[Marmite]

Veritas, don't forget about the possibility of two single phase faults to earth affecting different phases. Earth grids are invariably conservatively designed. The bulk of the cost is in installation, rather than going from 50mm to 70mm Cu. The maximum foreseeable current should be used. In Australia, from memory, AS2067 gives a minimum size of 35mm because the earthing needs to have some physical strength to give it a level of integrity. I wouldn't want to see you design an earth grid for a 25A fault current otherwise it is likely to disappear in a puff of smoke!
In terms of EPR the transformer winding doesn't matter. What if you get a Tx bushing fault, or a cable box fault if it's cable connected? Are there 132kv arresters, station post insulators etc? Any of them could flashover/conduct to earth raising the earth potential.
I would first calculate your EPR without factoring in current returning via the transmission line earth wire. If it is within limits then you need go no further. If it's out of limits then bring in the earth wire to the calculation, and any other fortuitous earth return paths such as the sheaths of distribution cables connected to the same sub.

Regards
Marmite

 

[veritas]

Marmite - your point regarding bushing or insulator flashovers is a valid one I did not take into account. Thanks. In essence then one could have a 132kV EF from the 132kV conductor to the earthgrid with the EF current traveling down the earthgrid to the source thereby creating an EPR.

I was not going to size the earthgrids based on 25A as the 11kV fault current with a generator EF is in the order of kA (have not done the detailed calcs yet). Also, this is a remote unmanned site so I cannot foresee many issues with transfer potentials when tying the grids together. As you say there is the benefit of reduced grid resistance.

The utility has advised that a 50:50 split be used for 132kV EF current to ground and via the OHEW. This means about 13kA going to ground.

I can see the problem of transfer potentials between the two grids if they're not bonded together. There are services, control and power cables running between them and so I agree it's best to tie the earthgrids together.

Thanks again for the insights marmite.

Regards.

 

[Sharpie8]

Veritas, there are several things to consider:
- does the local supply authority allow the two grids to be connected together;
- what is the soil resistivity in the area? If it is high then joining the two grids together may not be best.
- are the earth grids existing? If so you can test to understand the current configuration and how the currents split, identify touch hazards, etc.

Then there is the question of what standard are you designing it too? From your previous posts I believe you are in Australia, so are you designing to AS2067, or Ena eg0 (risk based approach), Ena eg1, local authority requirements/design requirements.

I don't think you can simply ask the question of if I should bond the two grids together, get an answer and run with it. You really need to model it in some software to understand what the ramifications of each decision are. There have been several earthing related fatalities in recent years so this is an area that needs thorough investigation.

 

[veritas]

Sharpie8

Thanks for the pointers.

Yes, it's and Australian design to AS 2067 and ENA EG1-2006 and drawing upon IEEE 80:2000. This is a greenfields project so no existing earthmats. Am using a two layer soil resistivity model. Soil tests have revealed resistivity of 10ohm.meters up to 50m and 8.49 ohm.meters further down.

PTW Groundmat is being used to do the modelling. So far no problem with step and touch. The power station is a temporary addition as the utlity supply will not be ready in time.

I am not responsible for the detailed design but have a keen interest in earthing and am assisting.