Testing Large Earth Systems

Introduction

Large earth systems are an important part of the protection of the electricity supply network. They ensure that fault current will enable protective devices to operate correctly. A substation must have a low earth resistance to reduce excessive voltages developing during a fault which could endanger safety of nearby people or animals.

When installing an earth system the resistivity of the nearby ground should be measured. Inaccurate resistivity tests can lead to unnecessary cost in the design of the system.

After installation it is vital to check that the earth meets the design criteria and it should also be measured periodically to ensure corrosion or changes in the soil's resistivity do not have an adverse affect. Earth networks may not appear faulty until a fault occurs and a dangerous situation arises.

In order to obtain a low enough value of earth resistance, earth systems may consist of an earth mat covering a large area or many interconnected rods. Suitable test techniques must be used for large systems to ensure valid readings are obtained. This is unlike a small single electrode (for example, a lightning conductor or domestic earth) which can be simple to test.

Unlike most electrical testing a single test is not sufficient to quantify the result. Despite the low conductivity of soil and rock, large volumes surrounding the earth electrode enable a low resistance to earth to be obtained. This volume has to be taken into account when testing.

 

Why is a Low Earth Required?

A good path to ground is essential so that the earth system can operate as required. This will enable protective devices to operate and reduce ground potential rises (GPRs).

The GPR can be calculated using Ohms Law from the prospective short circuit current and the earth resistance. This gives a figure for the maximum voltage difference between a remote earth and the earth system. Typically the short circuit current will be one to ten thousand Amps so even with a 0,2 W earth resistance the GPR will be 200 V to 2 kV. The maximum allowed value is typically 430 V to 650 V according to CCITT (telecoms) standards.

Step and touch potentials can also be calculated to illustrate the danger of shock to nearby people during a fault. Touch potential is the possible fault voltage between a foot and a point that is connected to a part of the earth system (for example, a substation perimeter fence). Similarly, step voltage is the potential difference across a nearby person's feet (1 m apart) during a fault.

 

Fig. 1: Step and Touch Voltages

 

If the GPR is too large a danger also exists during a fault, from electric shock via incoming conductors, for example telephone lines. The telephone line will be earthed at some remote point. The fault can cause a "transfer voltage" equivalent to the GPR.

The step voltage varies within the resistance area of the earth system. The volume of earth near the earth electrode has the most effect on the earth resistance and so this is where the most dangerous step voltages are usually found.

Fig. 2: Variation of Resistance and Potential from Earth Electrode

 

Testing Earth Systems

To test an earth system, a current is injected between a temporary, remote electrode and the earth system. A second temporary electrode is used to measure the resultant potential created by the test current. From the voltage and current values the resistance can be calculated.

 

Fall of Potential Testing

The test connections for a small electrode system are shown in fig. 3. For single electrode earths, such as domestic earths and lightning conductors, the influence on the surrounding soil is limited and current test spikes can be quite close (typically 10 to 20 m) to the electrode under test. It is usually quite easy to find a flat portion of the earth resistance curve which should be close to the resistance of the electrode.

Fig. 3: Test connections for Fall of Potential Test

 

Fig.4 shows an example of a small earth system with a test spike at 50 m. Using a potential spike distance of between 10 and 40 m, a reading close to the earth resistance will be measured. At distances less than 10 m the influence of the electrode under test will affect the measurement. Above 40 m the "resistance area" of the current spike will give a higher measured value than expected.

Fig.4: Theoretical Resistance Characteristic of a Single Earth Electrode

Testing several points or drawing up a curve will help the understanding of the area around the electrode. It is always best to check results by using a different direction or a longer distance to the test spikes. This will help to eliminate errors caused by nearby buried conductors and other parts of the electrical system interfering with the results.

 

Large Earth Systems

The physically large areas used by earth systems such as those in substations and power stations result in large "resistance areas" and consequently large distances to the test spikes. This can typically give a value of resistance to earth of less than 0.5 W allowing a good path for the large prospective fault current.

Ideally the distance to the current electrode should be ten times the maximum dimension of the earth system. For a single, 2 m long electrode this is not usually a problem with a remote test spike at 20 m. However, this may be impractical for a substation with a 100 m square earth mat. A current electrode is required approximately 1 km from the site. In cases like these a measuring technique such as the slope method can be used. This reduces the length of cable runs and is less likely to overlap with other local earth systems which may interfere with the result.

5: Theoretical Resistance Characteristic of a Large Earth System

 

Testing Large Systems

Up the Resolution

The specification of the MEGGER DET2/2 makes it ideal for measuring earth systems including substation earth mats, power station earthing and communication systems. When measuring earth resistance of less than 1 W, the 1 mW resolution allows real readings to be made without instrument errors overwhelming the results.

Using either the Fall of Potential Method or the Slope Method on large systems, means that small differences between low readings are required. The extra digit of resolution makes these variations more accurate and suitable for use with the published tables.

 

The Slope Method

The physically large areas used by earth systems such as those in substations and power stations result in large "resistance areas" and consequently large distances to the test spikes.

The slope method enables measurement of a large earth system without finding the flat portion of the characteristic curve. This can reduce the test distances and, in addition, the electrical centre of the earth system is not required for measurement, few calculations are necessary and the result can easily be checked giving added confidence to the test.

The slope method involves taking 3 readings at 20%, 40% and 60% to the current spike distance. The differences between these readings are used to fit to a mathematical model of the resistance characteristic. The coefficient of slope, m is calculated from;

A table of values of m versus "actual distance" is published in appropriate user guides. This value can then be retested to see if it fits the model.

As with all earth testing, it is best to check the results by plotting the full characteristic and repeating the test using a different direction for the test spikes or a larger distance to the current electrode.

Noise Interference

High noise interference rejection allows earth resistance readings to be made even in the presence of induced noise. A small test signal has to be retrieved from a much larger total signal.

To remove the effect of noise from an earth test a frequency of 128 Hz is frequently used. This is close enough to line frequency to give a result that can be used to make calculations of earth fault current. This frequency avoids harmonics of the standard line frequencies to allow filtering of the test signal. A filter can then remove the 50 or 60 Hz interference from the total signal.

Many earth testers can only reject noise of a single frequency. This may be acceptable in a laboratory but is inadequate for most real situations. Electrical networks contain noise consisting of the fundamental frequency of the supply and it's harmonics plus high frequency noise from switching etc. and induced signals from other sources. This type of interference can cause significant measurement errors without alerting the user. The instrument cannot reject the noise even though it is insufficient to trigger a high noise indicator.

Sometimes electrical noise may be short term and testing can be delayed until the noise has decreased, for example, a passing train when testing a railway system. However in most cases background noise cannot be removed and so a suitable instrument specification is required.

The MEGGER DET2/2 uses a sophisticated filtering system that can reject more noise than any other earth tester we have tested. Test frequency adjustment and selectable levels of noise filtering also help to remove stray noise that could affect the reading. A high current range increases the test signal strength in comparison with the "noise".

In extreme cases it may still be necessary to carry out the test when the noise has decreased, however, the DET2/2 can keep going at higher resolutions long after other earth testers have "given up".

 

Resistivity Surveys

Soil resistivity surveys allow data to be collected for use in earthing system design. The resistivity value is the fundamental value used for design calculations for the physical size of the earthing system and for the ground potential rises to be calculated. For this reason it is very important to attain maximum accuracy in the measured value.

The standard technique for resistivity measurement is pictured below to give a resistance value, R. With a spacing of 'a', the test can measure the average soil resistivity between P1 and P2 down to a depth equal to 'a'. Using different spike spacings it is therefore possible to test at different depths.

 

Fig. 6: Resistivity Measurement

 

The soil resistivity is then calculated from;

A high display resolution is essential for accurate resistivity surveys to be taken to prepare data for earthing system design. When testing with large test spike distances, for large electrode sites or deep surveys, resistance readings are small. To obtain reliable resistivity measurements it is essential to use an instrument with suitable accuracy at low values.

Example; With a test spike distance of 30 m in clay soil with a resistivity of 1200 Wcm. The resistance reading would be 0.064 W.

A tester with a 10 m W resolution and 3 digit error could give measurement errors of 50% of reading in this example. This could lead to either a dangerous earth with too high resistance or being over engineered leading to excessive cost.

A 1 m W resolution and 3 digit accuracy reduces this error to less than 5% of reading.

 

Conclusions

The latest generation of digital earth testers greatly simplify the testing of electrical earth systems. However, care is still needed interpreting the results. Error indicators can alert the user to misconnected leads or conditions that may lead to an invalid reading but simply taking one reading is not sufficient to measure the resistance of any earth electrode.

It is always best to repeat an earth test using a different direction or distance to verify the results. This may remove any errors from hidden differences in the soil and increase confidence in the results.

When selecting an earth tester, ensure that the resolution and accuracy are suitable for the application. Instrument errors can lead to unnecessary expense in the design or maintenance of earth systems or, worse still, unsafe installations.

Use a DET2/2 if you are testing low values (<1 W), or in the presence of induced noise. A high level of noise filtering is required for accurate results in real life situations.

MEGGER Digital Earth Testers

Model

DET62D

DET5/4

DET2/2
Terminals 3 4 4
Resolution 0.1 W 0.01 W 0.001 W
Maximum 2 k W 20 k W 20 k W
Accuracy 2% + 3D 2% + 3D 0,5% + 2D
Ranges 2 4 5
Noise 40 V 40 V 40 V
Broadband Noise Rejection No No Yes
Power 6 x AA 6 x AA Rechargeable or rechargeable