Ground Grid Resistance - Hydro Electric Plant

[Elmir]

On an upcoming project we are required to measure the ground grid impedance of a ground grid installation for an outdoor 230kV substation connected with a hydro electric plant. The ground grids longest dimension is a little over 100 meters. A penstock approximately 600 meters long exists which is connected to the plant. We are required to measure the ground impedance with and without the penstock connected.

With the penstock not connected we will place the remote current electrode 1000 meters away from the substation and run a fall-of-potential test using frequency selective equipment capable of measuring under and over the power system frequency.

To my understanding the ground grid was modelled using ETAP, therefore no complex analysis exists such as would be obtained using CDEGS.

With the penstock connected would I have to consider the ground grid being 100 meters plus the 600 meter penstock?

If this is the case I would likely have to enquire about using the tranmission line as the current return path and place the remote current electrode at least 6km away. Your thoughs and comments are appreciated.

 

[jghrist]

I have heard of this being done, but don't know of any details. I suppose that you would use two phase wires, one to connect the remote current probe and one for the potential probe. A problem I see would be placing the potential probe without being influenced by tower grounds. Would the shield wire be isolated from the grid? Another problem might be mutual coupling of the potential and current probe wires.

 

[Elmir]

I am not certain on the length of shield wire at the moment but I do know it only extends a few towers out from the substation. This is regular practice for BC Hydro tranmission lines.

At this point I would have to have discussion with the engineer and utility to determine feasibility of isolating the shield wire for testing if interference is a concern. My past readings on tranmission station grounding have been performing the test with the shield wire connected since this influences the ground grid impedance.

The terrain is difficult and only practically allows for following the transmission line since the station is in a mountainous region.

 

[jghrist]

The tower grounds will not affect the measurements if there is no shield wire connecting the grounds together.

My concern with mutual coupling is between the phase wire used for the current lead and the one used for the potential lead. See Influence of Inductive Coupling Betwen Leads on Ground Impedance Measurements Using the Fall-of-Potential Method, L. Ma and F.P. Dawalibi, IEEE Transactions on Power Delivery, Vol. 16, No. 4, October 2001, pp 739-743. Large errors are possible with large grids or low soil resistivity.

 

[eamonkerrigan]

Check out the Omicron website. They have an interesting test method utilizing the transmission phase wires for current path. I had this recently demonstrated. It helps to have grounding switches. Anyways, they have a video on thier website which may help at least visualize the test. I will warn you that the demo I had done was not successful as the vendor forgot to account for mutual inductance (see jghrist comment).

 

[jghrist]

The Omicron method is interesting. A description is at

http://www.omicron.at/fileadmin/user_upload/files/pdf/en/2006-Earthgrid-Impedance.pdf.

With this method, the voltage probe is not connected with a phase wire and the potential wire will be 90° from the current wire.

 

[Elmir]

We are using the same Omicron equipment for our testing. Due to the terrain we will likely not be able to accomplish 90 degrees between the potential and current wires.

We could measure the mutual coupling of the overhead line with the Omicron equipment and take this into account to evaluate the ground grid impedance.

 

[Elmir]

Does anyone have any comments regarding how far out to take the remote current electrode when performing the fall of potential test....Do I treat the ground grid as 100m or 700m when connected to the pen stock?

 

[Zogzog]

The current probe for a FOp test would need to be thousands of feet for a system like this, you are better off using the slope method or intersecting curves method.

 

[jghrist]

If you are using the Omicron method with the transmission line, the current probe is the remote substation ground grid, miles away. You need to keep moving the potential probe out and plotting points until there is very little voltage difference between successive spacings.