Resistvity-Zone of Measurement

[Meinkaun]

Hi,...

I am using ETAP for grounding grid simulation. Very shortly, we are arranging for the resistivity tests and I have done some reading to understand these IEEE approved methods. Most of the commercially available softwares including ETAP, SKM, WinIGS have two options:

1) Two Layer Model
2) Test Values interms of probe spacing etc.(wenner method etc.)

What I am having difficulty in understanding is that the soil layer model even if it is single layer or multi layer: How far from the vicinity of the substation location, the uniformity of this soil layer is required.
for example, if I have a 20mX20m size substaion: Does this soil layer uniformity should exist 20 m.... in either direction.... or how does this work.

Thanks

 

[jghrist]

The question is not how far soil uniformity should exist. The soil around the substation is either uniform or it isn't. You don't choose a substation site based on uniformity of the soil resistivity.

The question is, how far apart do your probes have to be to adequately represent the soil for purposes of calculating grid resistance, step-voltage, and touch-voltage. Increasing probe spacing does two things. It increases the area covered by the measurements and it measures resistivity at different depths. You have to measure at several different probe spacings to determine the soil model at different depths. You have to have a maximum probe spacing that covers the zone of influence of the ground grid. If you make the maximum probe spacing 4-5 times the maximum diagonal of the grid, you will ensure that your measurements extend throughout the zone of influence of the ground grid, both horizontally and vertically.

 

[Meinkaun]

Jghrist,

I noticed a lot of contribution from you on grounding issues on this forum and thanks for responding to my question. A word of advise from a GURU like you would go a long way.

What I understood from your response is that The uniformity in terms of depth only matters up to Zone of influence.

When you talk about zone of influence; what is this value?

I am picturing a real time situation, which I am planning to do as a modelling scenario on one of the substations and I would appreciate, if you could correct me if I am wrong.

A 80 ft by 20 ft substation in a rocky area. I have not done any tests as it is possibly all rock. Rather measuring rock resistivity, which is kind of impractical as it is not easy to drill probes in rocks(As told by Testing contractor) I assumed a value of 1000 ohm -m.

I decided to pursue a two layer model, which is a default option in most of the simulation softwares. I have assumed rock resistivity of 10,000 ohm-meters(Typical IEEE Std. value) and excavate up to 8 ft of the site and fill with organic soil with resistivity 0f 100 ohm-m. It gives me a resistance of 24 ohm.

Technically, this is a CUBE 80*40*8ft of organic soil and besides it is rock.

Do you think this physical arrangment would be acceptable as two layer model..

My apologies for such a long question, I just wanted to make sense that it should make sense.

thanks

 

[jghrist]

The zone of influence if the area where most of the voltage drop to remote earth occurs. Beyond the zone of influence, voltages are close to that of remote earth and the voltage gradient is small. You hear reference to this in making grid resistance measurements, where if the zones of influence of the grid and the current probe overlap in a fall-of-potential measurement, there will be no flattening of the potential plot vs distance.

I would discuss the measurement problems with a test instrument manufacturer such as AVO (Megger) or AEMC. You do not have to drill into the rock to test the rock resistivity. If there is soil over the rock, a probe in the soil will inject current through the rock. Even if it is all rock, you can get measurements by laying a mesh probe over the rock. I haven't done this, but I have seen reference to this method.

You should measure the soil resistivity. Assuming 10,000 ohm-m of resistivity for an infinite depth could be over conservative. There may be a water table not too far down.

Excavating 8 ft of rock will be very expensive, so you want to have as accurate a model as possible to assure that it will achieve the desired results before going to this expense. If you only put the low resistivity soil in the area of the substation, the fault current flowing out of the grid to the source will have to flow into the rock after it get outside the substation or below the soil. I would not expect adding the soil to greatly reduce the grid resistance. It may reduce the touch-voltages to an acceptable level, however.

A standard two-layer soil model assumes an infinite extent of each layer horizontally, so it will not be an accurate model, particularly around the perimeter, when the low resistivity layer is only over the grid. Some programs, such as SES CDEGS MALT, will allow modelling a finite volume of different resistivity soil.

 

[cherry2000]

MeinKaun
Based on IEEE81, I have developed a excel based program, where the 4 pin Wenner method readings can be converted to a two layer soil model.

 

[spredrag]

Cherry2000,

Please can you send us your excel file or send us some information how to convert the 4 pin Wenner method to 2-layer soil model. (We received information from the civil contractor with 6 probe spacings from 12 different points on the site)

Thanks.