Wind Farm grounding problems

[JLuc]

Hi all,

we are working on design and construction of ground grids for a wind farm.

For each 1.5 MW generator (80 total), GE is specifying ground grid resistance of 2 to 10 ohms.

With the data we have:
-ground resistivity : 500 to 1000 ohms, average
-projected ground grid area: 285 m^2, (about 3060 square feet)

First calculation (from IEEE 80-2000, formula 50 & 51):
we get ground grid resistance of 20 to 40 ohms (perfect condition), which is not bad for the type of soil we have, but still way over GE specification (2 to 10 ohms).

Two solutions came up:

1. tie all grids togheter with 3/0 AWG bare copper wire to lower the resistivity of each grid, we are confident this would work, but this solution is VERY expensive

2. Here's my question...: when the 34.5 kV power cables will be installed, if all concentric neutrals are tied at both end to each ground grid, this will have the effect of tying all grids togheter,at no extra cost, and will probably lower the resistance to satisfy GE specification.

I would like to have your opinions about this option.

Note: for grid construction, we plan to drill 3" holes in the roc, and backfill 3/4" 10' copper rods with Erico GEM. Grid conductor will be 4/0 AWG bare copper.

Thank you.

 

[seap]

While connecting all turbine ground grids by the concentric neutrals of the MV power cables will lower ground resistance for power frequency faults, and in most cases decreases GPR and touch and step potentials to tolerable levels, it has little effect on the grid resistance at lightning frequencies upon which the 10 Ohms limit value is based. Usually the turbine supplier and Client specify that the 10 Ohms value must be met by each base independently (i.e. with concentric neutrals and counterpoise earths disconnected from remote earths).

In my experience, the concrete turbine base and associated re-bar have a significant effect on the expected ground resistance and values of > 10 Ohms are rare. However if values > 10 Ohms are encountered, the best way to reduce the measured value is to lay radial spurs of earthing conductor of approx 50m length in various directions from the base or drive additional rods.

Incidentally, in Europe it is common wind farm practice to lay a 50mmsq bare copper conductor with the MV cables as well as earthing the concentric neutrals of the MV cables at both ends to the turbine bases.

 

[jbond]

I've designed the earthing systems for lots of small off-grid powerstations where the national regulations require the grid resistance to be no greater than 1 ohm. Unfortunately these might be small power systems with a total station capacity of less than 600kW (I'm trying to convey the sense that the budget is not big). The soil resistivity at one site measured between 1000ohms/m to 2000ohms/m depending on the depth (multiple layer model).

Fortunately we were able to get station grid resistance down to 0.2ohms on a small budget.

The water table isn't too deep under the ground at some of these sites, so I got a contractor to bore down to the water table at 4 locations on the site (only about 14 metres below the soil surface) and then we installed ground rods and back filled with bentonite. The spec I wrote called for the drilling to proceed 1 meter past the water table to ensure that contact would still be present in case the water table varied with seasons or over time.

Can you try something like this at your windfarm site?

And yes, the concrete and reo should help to drop the grid resistance further, it's usually not modelled in your calcs or software so it's a bonus. Out of curiousity, how do you bond your earthing to your reo?

Additionally, I think I've heard before that lightning strikes can cause great heating effects in your reo etc if it's part of your earthing, and can lead to cracking in your concrete - has anyone heard of that or can shed some light on that concept?

 

[7anoter4]

If the generator is connected through an isolating transformer the generator grounding current does not leave the grounding grid.
So, the generator grounding resistance is only between generator and grounding grid.
This is my opinion.

 

[Sungate]

In our wind farm project we are thinking to disconnect concentric neutrals of the MV cables at one end in purpose to cut gronding loop as we have star-star pad-mount turbine transformer. MV system is ungrounded so that could prevent neutral floating as well . I would not consider gronding connection link between WT towers in any case.