To know whether a grid performs in practice, IEEE 81 proposes injecting a known current between the grid and a remote earth, measuring the surface potential gradients and scaling the result to the real fault current. It describes six types of touch voltage, four ways of measuring and the problems that falsify the result.
Short answer. Calculating the step and touch voltages of a grid in design is a study; measuring them on an installed grid is a field test. IEEE Std 81-2012 (clause 9) proposes the most practical way: inject a known current between the grid and a remote earth, measure the potential gradients at the surface and scale the results to the real fault current. The most accurate test, with a real fault, is almost never done.
Why they are measured
According to clause 9.1, local potential differences are measured for two purposes. One is to check the step and touch voltages of an existing substation against the values the human body tolerates. The other is to confirm the design calculations of a new substation. The real values can differ from the calculated ones because of the approximations of the model.
A grid is designed to limit those voltages to tolerable values. But the soundest assurance that it meets its purpose is to measure them: a known current is injected from a remote electrode, the resulting gradients are measured and scaled up to the substation fault current.
Types of voltage
Clause 9.2 distinguishes two voltages. Step voltage is the surface potential difference between two points 1 m apart. Touch voltage is the difference between the grid potential (GPR) and that of the surface where the person stands while touching a grounded structure. Touch voltage has six variants:
| Type | Where it is measured | Where it is usually highest |
|---|---|---|
| Structure | Between a structure or metallic object in the yard and the ground at 1 m | At structures near the edge of the grid |
| Mesh | Between the grid and the ground within the footprint of a grid mesh | Near the center of a perimeter mesh |
| Fence | Between a metal fence and the ground at 1 m | At a grounded fence, outside the corner farthest from the center; at an isolated fence, also inside near the grid |
| Gate | Between a metal gate and the ground at 1 m, at any swing position | Points less than 4 m from the closed position must be checked, because many gates swing in a 3 m arc |
| Substation transferred | Between the grid and the ground at 1 m from an outside metallic object bonded to the grid | The sheath of power cables on a pole, the external valve of a water pipe or rails |
| Remote transferred | Between a remotely bonded metallic object and the ground at 1 m, near the substation | A water faucet in a neighboring house bonded to the multigrounded neutral, or the shield of a telephone cable |
Four ways of measuring
Staged fault. With the full current of a phase-to-ground fault, the most accurate data are obtained. The guide states that it is rarely done, unless there are other weighty reasons, such as determining circuit parameters, equipment behavior or protection characteristics. It demands special care and advance preparation.
Current injection (low-voltage fault). A power-frequency current of 100 A to 200 A is circulated between the grid and a remote grid or electrode. Meanwhile the current distribution, the GPR and the voltage gradients are measured. Sometimes the measurement includes the current through a 1000 Ω or 500 Ω resistor that simulates the human body. The values are multiplied by the ratio between the substation fault current and the test current. Normally a line is de-energized and its phases carry the current from a source. It can be a mobile substation, a temporarily installed transformer, a gasoline-driven arc welder or a portable generator. With the last two, the regulator is set for a few hundred amperes at a frequency slightly different from power frequency. The guide also describes a variant with reversal of the current direction to eliminate stray power-frequency currents and harmonics, which allows testing without de-energizing the whole substation.
Conventional ground meter. The available ground meters work with ranges of 0 V to 50 V and 0 mA to 250 mA. Their frequency ranges from 40 Hz to 500 Hz, excluding 50 Hz and 60 Hz, and they measure from 20 mΩ to 300 kΩ. They are little used for step and touch voltages, although it is possible if the equipment can inject enough current at a frequency other than power frequency. The equipment measures the impedance between the grid and a point in the yard (touch voltage per ampere). Multiplied by the estimated fault current, it gives the touch voltage at that point. Step voltage is obtained by subtracting two touch voltages measured 1 m apart. If the current probe is too close to the grid, the measured voltages come out lower than the real ones.
Computer-based multimeter. The same equipment that measures impedance measures voltages with the probes at the points of interest, filters the noise and delivers six voltages normalized to the fault current that was entered. It compensates for the proximity of the return electrode and for the impedance of the other grounds connected to the grid. It is described in clause 9.4.4.
What goes wrong in the measurement
Clause 9.5 lists the most common problems:
- A conductive shoe to rest on the surface will probably fail if it does not achieve the desired contact resistance.
- The local potential differences are small, a fraction of the GPR. The instrument needs resolution better than 1 mΩ and good cancellation of 60 Hz noise.
- It is hard to know where the highest potentials will be, because only a few points can be tested. Grid drawings or a cable locator help.
- Seasonal changes in moisture have an effect: the highest potentials normally occur when the surface resistivity is at its maximum. Hilly areas without vegetation and with sandy or gravel subsoil are the ones that change most.
- The result requires adjustment if the injected current produces a different split in interconnected shield wires or neutrals than a real fault.
- Asphalt or concrete surfaces are difficult to probe; sometimes it is possible to measure in a nearby area at a similar distance from the buried conductors.
- Errors can be introduced by potential differences along the grid, especially near where the current generator is connected.
- The current probe must be far from the grid; for greater accuracy, at least five times its largest dimension.
The stone in the yard also has an influence. Gravel deteriorates when clay from the subsoil invades it, weeds grow or it becomes contaminated with fines. A test of the surface material resistivity helps set the tolerable limits. To assess safety, the measured voltages are compared with the IEEE 80 limits, which depend on the tolerable body current, the resistance under the feet and the fault duration.
If you want to know the effects of a high step voltage on a grid in service, read the guide on high step and touch voltages in the grid. The step and touch voltage guide explains what they are and why they matter. For the map of the installed grid and its integrity, see Potential map and integrity of the installed grid. Before injecting current, review the safety risks of measuring; and for the instrument that also measures these voltages, the computer-assisted ground multimeter.
We measure your grounding system
At Tiza Group we measure step, touch and transferred voltages in substations and plants, in Colombia, or elsewhere on request, and we deliver the result scaled and compared with the tolerable limits. See the testing and commissioning service or write to us through contact. In the quotation we recommend the test and the equipment your case needs.
Related guides
- Potential map and integrity of the installed grid
- Tolerable body current according to IEEE 80
- Substation gravel resistivity: how it is measured
- Safety when measuring a grounding system: real risks
- Step and touch voltage: difference and IEEE 80
References
- IEEE Std 81-2012, IEEE Guide for Measuring Earth Resistivity, Ground Impedance, and Earth Surface Potentials of a Grounding System: clauses 9.1 to 9.5 (purpose, types, procedure, methods and measurement problems).
- IEEE Std 80, IEEE Guide for Safety in AC Substation Grounding.
Frequently asked questions
How do you go from the measured voltage to the voltage during a real fault?
By scaling. The voltage measured with a known test current is multiplied by the ratio between the fault current that would enter the grounding system and the test current. This gives the step, touch and transferred voltages that would occur with the fault, which are then compared with the tolerable limits of IEEE 80.
What test current is used in the injection?
IEEE 81 describes a power-frequency current in the range of 100 A to 200 A between the grid and a remote earth. The higher the test current compared with the noise, the shorter the measurement can be.
Can step voltage be obtained without measuring it directly?
Yes. Step voltage can be obtained by subtracting two touch voltage measurements taken at points 1 m apart. With a conventional ground meter it is done like this: the impedance between the grid and each point is measured, and multiplied by the estimated fault current.