It is important that the resistivity is assessed as accurately as possible, since the value of the resistance of the electrode is directly proportional to the soil resistivity. If the incorrect value of soil resistivity is used at the design stage, the measured impedance of the earthing system may prove to be significantly different to that planned. This could, in turn, have serious financial consequences.
The test is traditionally carried out using a four-terminal earth tester. Four spikes are driven into the ground as shown in the diagram, spaced a distance of “a” metres apart. The depth to which each spike is driven should not exceed “a” divided by 20 and is not normally greater
than 0.3 metres. The outer two spikes should be connected to the current terminals C1 and C2 of the instrument, the inner spikes to the potential terminals P1 and P2.
It is important to ensure that the test spikes are not inserted in line with buried metal pipes or cables, as these will introduce measurement errors.
If “R” is the instrument resistance reading in Ohms, for a separation “a” metres, then the apparent resistivity is given by the following formula:
Resistivity = 2 x ã x R x a Ohm-metres.
The term “apparent resistivity” is used since the above formula assumes that the soil is uniform to a depth “a” metres below the centre point of the measurement traverse. We are able to obtain information about the actual soil layering by taking a series of readings, with “a” being increased by 1 m steps up to 6 m separation, then by 6 m steps up to typically 30 m separation. For very large area sites, especially where there is rock beneath, readings may be advisable at 50 m, 80 m and even 100 m spike separation. The test instrument used should be sufficiently accurate to measure quite small resistance values at these large spacings - in the order of 0.01Ω to 0.002Ω. The measurements should preferably be made in an area of reasonably undisturbed soil. Typically the lower values of “a” will give high values of soil resistivity because they are heavily influenced by the surface soil which normally drains or has its water content reduced by sun and/or wind. As the distance “a” increases, the apparent resistivity would normally decrease, unless there is underlying rock.
A curve of resistivity against separation should be drawn during the measurement exercise. This will provide information on the general structure of the soil in the locality, identify rogue readings and help in deciding how many measurements are required. If there are large fluctuations in measured values, then it is likely that the soil conditions are variable, the ground has been made up, or there are buried pipes in the area. In all cases, measurements should be taken on a number of traverses across the site. Some of these traverses should be at right angles to one another to enable any interference from nearby electricity cables to be identified.
Some examples of soil resistivity curves are shown in Figure 6-7 and Figure 6-8. In Figure 6-7, a number of measurements have been taken at the site and there are variations between them. The apparent resistivity value is higher at short spacings and then falls into a reasonably narrow, uniform band. Computer analysis produces a two layer model where the surface layer is 0.2 m thick and has a resistivity of 126 Ohm-metres. The underlying material has a value (biased towards the higher readings) of 47 Ohm-metres.
Figure 6-7 Apparent soil resistivity plotted against test spike separation - relatively uniform soil
For practical purposes one would assume a uniform soil of 47 Ohm-metres, since the value of the surface layer will change throughout the year. In the second example (Figure 6-8), the readings are much more difficult to interpret and analysis via computer software produces a three layer model. The middle layer has a low resistivity, so vertical rods or horizontal electrodes installed at greater depth than normal would be used. The actual readings are shown to be either side of an average computer model and typifies the variation expected on different traverses across the same site. The average three layer model would normally be used for earthing calculations.
Test spikes should not normally be installed within 5 metres or so of an electricity substation, unless suitable precautions are taken. The buried cables there will influence the readings and should an earth fault occur whilst testing is taking place, the potential gradient near the substation may be sufficient to introduce a risk of electric shock to those carrying out the test.
Figure 6-8 Apparent soil resistivity plotted against test spike separation - three layer soil
The method of soil resistivity measurement described above is the Wenner method, using spikes which are equidistantly spaced. There are other methods available for use in more difficult locations. These include the Schlumberger technique where the distance between the instrument and each current spike and each voltage spike is the same, but that between the voltage and current spikes is different. This is illustrated below:
Software is also available which can enable the soil resistivity to be calculated when the spacing along the traverse is arbitrary. This may enable soil resistivity readings to be taken when there are physical obstructions (roads, pavements, concreted areas etc.) preventing use of the Wenner method. Finally, another method of determining the soil resistivity involves measuring the resistance obtained at different depths as an earth electrode is driven into the ground (the method of taking the measurement, but not how to interpret the readings is covered in chapter 13). The measurements are repeated at a number of locations around the substation and the average values used to determine the soil resistivity and layering. Because of localised effects, this method is not generally as accurate as the Wenner and other techniques, but may be the only one method available is some built up areas.