Are we really testing R, Z or something else?
In theory, grounding system testing is very simple. One simply applies a voltage between the system under test and a remote earth electrode, measure the current, measure the voltage ‘across’ the system under test, and divide it by the current.
In practice grounding system testing is somewhat more complicated. Questions often arise regarding the practicalities of testing, including:
- How do you establish a suitable remote earth?
- What type of voltage do you apply?
- How much current flow do you need?
- Where & how specifically do you measure the voltage?
- How do you measure both reliably?
- How do you detect and reject (or correct for) electrical noise, both AC & DC?
These are all important questions, and whilst many have reasonable answers, one question remains in dispute: If you test with an applied voltage that has both DC & AC characteristics, do you measure a resistance or an impedance?
The IEEE committee revising Std 81, amongst others, have wrestled with this question, and currently, there is no consensus. To answer the question “Are we really testing R, Z or something else?” the authors have investigated and tested a range of scenarios and instruments and sought to be definitive. This article presents their methods, results, and conclusions.
Why are we asking this question and what are we testing?
For an grounding system to be effective, its performance needs to be adequate for the power system attached to it and all the factors that affect its performance need to be considered. The performance of the grounding system is entirely dependent of the ground in which it is placed. The soil composition, existing infrastructure and mineral/moisture content all affect the apparent resistivity of the soil and thereby the performance of the grounding system. To assess the performance of an grounding system requires tests of parameters associated with or affecting that performance, including soil resistivity, EPR, resistance of earth grid, loop impedance, and continuity.
How are we testing?
To measure each of the aforementioned parameters current must be injected into the grounding system and/or the surrounding soil and the response measured. The injected current can be either a ‘Switched DC’ or an ‘AC current’ injection.
In both methods, the current is known, and voltage is measured. The question we are investigating here is “for a switched DC injection how should the measured voltage be interpreted”?
How do switched DC instruments work?
Switched DC instruments inject a known current in the form of a switched current square wave into the system. The voltage measurement is conducted at a specific part of the injected waveform, some in the middle and some at the end of the waveform. Measurements are taken at these points to remove the effect of test lead coupling. With a known current and a measured voltage, ohms law can be applied – (R=V/I) to deduce the measured resistance.
What happens when it gets complex?
Reactive elements in the system introduce complexity to an grounding system’s response. When measuring a complex system, the presence of reactive elements can affect the measured result. These reactive elements come from auxiliary paths such as overhead earth wires, buried conductors and cable screens.
How we showed the measurement can include errors
Using two switched DC instruments, we set up an equivalent circuit that would allow us to change real and reactive components. From this we could simulate how the measurement waveforms become distorted as the ratio of L to R increases and what the resistance tester records as the apparent resistance. The setup was measured with zero inductance initially to establish a control value. All other measurements with inductance were compared to the control value.
What happens when we have reactive elements?
Based on one theory the behaviour of reactive elements is just …
The reality is that high inductance reacts to the injected switched DC current such that the resulting voltage waveforms become distorted, as shown in Fig 3. The distortion in the measured voltage waveforms create error within the measured resistance.
The result is that the magnitude of the measurement is larger than the resistive component of the circuit and is not indicative of the system under test. If the reactive element is small in relation to the real resistance, there exists a plateau near the falling edge of the received signal. This plateau is the real magnitude and contains little to no error being uninfluenced by the reactive element. By using the plateau of the signal, the measurement can be effectively used to calculate the real resistive element.
How does impedance affect the results?
What we found was that the effects of inductance on measurement accuracy and reliability are dependent on the ratio of R to L, with the following features:
- Large R values negate low L values;
- Large L values dominate low R values;
- The L/R ratio does not imply the associated error;
- The error becomes significant when the millihenry range of inductance is reached; and
- The lower the frequency the more time for reactive elements to dissipate, allowing more time for a plateau to be reached.
Are we measuring an impedance with switched DC?
The short answer is no. Switched DC current injection measures the voltage at a certain point of the waveform and applies Ohm’s Law (R=V/I). It cannot measure a phase angle and thus cannot get a phase shift to deduce impedance.
The long answer is, it’s complicated. Due to the truncated measurement window technique used specifically by these testers, quantifying the amount of reactive information captured in that sample window is not trivial. In addition to this, the result is not a linear increase in error for a linear increase in inductance.
Does it matter?
Typically, switched DC measuring instruments are reliable for individual electrodes, groups of electrodes and small to medium sized grids with no or minimal interconnecting paths. When testing on large grids or grids with multiple auxiliary paths, the effect of reactive elements can be observed by changing the injected frequency and noting that the recorded resistance changes with frequency.
We ask again…. Are we really testing R, Z or something else?
These instruments are suitable for testing a resistance until they cross an implicit R/L ratio threshold. Past this threshold it appears that the inductive (L) component affects the measurement sufficiently such that substantial error can be observed. This measurement is relative to the circuit under test, but the reactive element cannot be deduced.
Through this investigation, it was found that 3 & 4 terminal switched-DC testers are always attempting to measure a resistance. When significant reactive elements are introduced into the circuit under test, the instrument ends up reading something else as it is trying to calculate resistance despite the measured waveform being distorted by the presence of the reactive elements. This implies that it is not an impedance being measured, the value that is calculated is only relevant to that specific circuit.
A way to check if there are reactive elements in the circuit is to change the frequency of the square wave. If the value doesn’t change, then it is dominantly resistive, and the measured estimate is valid. However, if the measured value changes significantly, there is enough of a reactive element in the circuit under test to affect the result and it must be considered an invalid measurement.
This article comes from the perspective of Safearth development engineers Matthew Lee and Steven Preedy who research, design, develop and manufacture our suite of Safearth grounding system test instruments. Our consultants regularly use our grounding testing equipment in the field in Australia and overseas.