Grounding in a test serves two purposes at once: it provides the reference against which the measurement is made, and it controls the potential of the parts of the arrangement that people can touch. Both purposes are affected by the same decisions, which is why the safety and the measurement arguments usually point the same way.

Why field grounding differs from laboratory practice

A laboratory provides a low-impedance earth reference that is common to the test set, the instrument and the test object, and it is not carrying other currents. A substation earth is shared, it carries the currents of the operating equipment, and currents from adjacent circuits flow through it during normal operation and during faults.

The consequence is that two points on the substation earth are not necessarily at the same potential. A test arrangement that connects to the earth at two different points therefore introduces a potential difference into the measurement that it did not intend.

The second difference is scale. A laboratory test object is metres from the earth reference; a field test may involve connections that run tens of metres, and the impedance of those connections becomes part of the measurement rather than a negligible detail.

Grounding the test object and the instrument

The last row is worth stating explicitly. Bonding the test earth to parts of the installation that were not intended to be part of the circuit creates a parallel path, and the current that flows in it appears in the measurement.

Return paths and their effect on readings

Every measurement involves a current that leaves the source and returns to it. The return path has an impedance, and any voltage developed along it appears in series with the quantity being measured.

The problem arises when the return path is shared. A return conductor that also carries the current of another circuit has a voltage developed along it by that other current, and the measurement includes it. The practical response is to keep the measurement circuit’s return path separate from shared conductors for the length of the test.

The same argument applies to the high-voltage lead. Its capacitance to earth and to nearby conductors forms part of the circuit being tested, particularly in a resonant arrangement where the lead capacitance affects the tuning point. The lead arrangement therefore belongs in the record as part of the measurement conditions.

Bonding of temporary connections

Temporary connections made for a test have to carry the test current without introducing a voltage drop that matters. The conductor size, the termination and the length all determine whether the connection is adequate.

The terminations deserve particular attention. A clamp that is convenient but makes poor contact introduces a resistance that appears in the measurement and, at test currents, a heating point. Where the connection carries the full test current, a bolted termination is the more reliable choice.

The bonding arrangement should be established before the connections to the test object are made, so that the reference exists while the circuit is assembled rather than being applied afterwards. Where the arrangement is changed during the test, the change should be recorded and, where it affects the measurement, the measurement should be repeated.

Interference caused by poor grounding

The symptoms are recognisable. Noise on a low-level measurement, an offset that changes when equipment elsewhere on the site is operated, and readings that vary between attempts without any change in the test object.

The diagnosis is usually a change of arrangement. Moving the earth reference, separating a shared return, or re-routing a lead that runs alongside a conductor carrying current will change the reading if the grounding is responsible, and will not change it if it is not.

There is also a safety symptom. Poor grounding leaves parts of the arrangement at a potential the team did not intend, and the assumption that a piece of equipment is earthed because it is connected to something metallic is not a control. The measurement system requirements that the arrangement has to satisfy are set out in IEC 60060-2.

Verifying the grounding arrangement before testing

The verification should establish that the intended reference point exists, that the connections to it are sound, and that the arrangement corresponds to the plan.

Where the earth resistance at the reference point is relevant to the measurement, it can be measured as part of the preparation. On a substation, the grid resistance is usually low enough that it is not the dominant term, but the continuity of the connection to it is always relevant.

The verification should be performed by someone other than the person who made the connections, in the same way that the test connections themselves are verified. That second check is the control that catches the connection made to a nearby earth point rather than the intended one.

Common mistakes that distort results

Using the nearest earth rather than the intended reference is the commonest. It is convenient, it appears to work, and it introduces a potential difference that varies with whatever the site is doing.

Running the measurement lead alongside a conductor carrying current is the second. The coupling produces a signal that appears in the measurement and that changes when the other circuit’s load changes.

Leaving a temporary bond in place from an earlier activity is the third. The bond provides a parallel path that was not planned, and its effect appears as an unexplained change in the reading once it is discovered and removed.

The fourth is a connection whose termination has not been checked. A clamp on a painted surface, a corroded bolted joint or a connector that is not fully engaged all introduce resistance that looks like a property of the test object.

Documenting the setup

The record should state where the earth reference was taken, which points were bonded and in what order, where the measuring instrument was referenced, and the routing of the high-voltage and return leads.

It should also state any earth connection that was made or removed for the test, and the state in which the earthing was left afterwards. On a substation, the earthing arrangements are safety-critical, and a temporary change that is not recorded is a change that may persist.

The record has a second use. Where a later measurement on the same equipment produces a different result, the setup documentation is what allows the difference to be attributed to the equipment or to the arrangement. The general test technique framework is published as IEC 60060-1 and the field testing requirements as IEC 60060-3; the installation requirements for earthing are set out in IEC 61936-1; safety guidance for the work is published by the UK Health and Safety Executive; instrument method documentation is published by suppliers such as Hioki; and practice is coordinated through CIGRE study committees. The earth test instrument range is grouped on the electrical resistance testing hub.