Earth Resistance Testing on Site: Fall-of-Potential, Clamp Meter and What the Reading Means

In short. An earth resistance reading is only useful if the method, the probe distances and the season are right. The fall-of-potential test with a proper earth tester, taken in the dry months with the three-reading check, is the reference method; clamp meters and two-pole tests are useful shortcuts with real limits. Reddy Generators tests generator earthing in Hyderabad, Telangana and Andhra Pradesh as part of installation, commissioning and AMC work, and this guide sets out how to do it safely and how to read the result. Pit construction and upkeep are covered in our earth pit types and maintenance guide.

Who this page is for: Maintenance electricians, consultants and facility managers in Hyderabad, Telangana and Andhra Pradesh who test, or commission the testing of, generator and building earth electrodes, and need to know how to take a reading that means something and how to judge it.

Safety first

What the instrument measures

Current injected into an electrode spreads into the soil through ever-larger shells of earth. Most of the resistance lies in the first few metres around the electrode; beyond that, the voltage barely changes. The test aims to measure the voltage between the electrode and that remote, flat region, divided by the injected current. Two auxiliary spikes are needed: a current spike (C) far enough away that its own resistance zone does not overlap the electrode’s, and a potential spike (P) placed in the flat zone between them. If the zones overlap there is no flat zone, and no position of P gives the true value.

Fall-of-potential, step by step

Why 62 %? For a small electrode in uniform soil, theory places the point at which the measured value equals the true resistance at about 61.8 % of the distance to the current spike. Real soil is rarely uniform, which is why some practitioners quote a range of roughly 62–68 % and why the three-reading check — not the percentage — is what makes a result trustworthy.

Spike positions for common current-spike distances

Worked example with readings

Illustrative case. Body pit 1 of a 250 kVA DG at a Hyderabad office campus: a 3 m GI pipe, tested in October, link opened with the set stopped and pit 2 still connected to the frame.

1. First attempt, D = 30 m: readings at 15.6, 18.6 and 21.6 m are 11.6, 12.1 and 12.9 Ω. Spread = (12.9 − 11.6) ÷ 12.1 ≈ 10.7 % — too wide; the curve is still rising, so the spikes are too close.
2. Second attempt, D = 50 m: readings at 26, 31 and 36 m are 12.2, 12.4 and 12.7 Ω. Spread = (12.7 − 12.2) ÷ 12.4 ≈ 4.0 %. Accept 12.4 Ω (the average, 12.43 Ω, is practically the same).
3. For confidence the tester walked the whole line at D = 50 m:

The plateau from about 50 % to 70 % is clear; the steep rise near the electrode and near the current spike is normal.

4. Seasonal correction from the site’s own records: last year this pit read 11.6 Ω in October and 21.0 Ω in April, a ratio of 21.0 ÷ 11.6 ≈ 1.81. Expected April value: 12.4 × 1.81 ≈ 22.4 Ω.
5. Pit 2 tested the same way reads 14.0 Ω, so about 25.3 Ω in April. The pair in parallel (ignoring interaction): about 6.6 Ω now, about 11.9 Ω in April.
6. Judging it: the site specification asks for 5 Ω or less for the body earth; the pair fails today and fails by more in April. Note what the reading does not mean: on this TN-fed site, faults on circuits are cleared through the protective conductors, not through the pits, so the installation is not unsafe by this number alone. Had the site been TT-fed with a 300 mA residual-current device, the safety condition RA × IΔn ≤ 50 V would allow up to 50 ÷ 0.3 ≈ 167 Ω. The specification exists for other reasons — stable neutral, surge and lightning performance, margin for ageing — and should still be met. Additional electrodes were planned using our earthing design calculation.

The three-electrode (ammeter) method

Before purpose-built testers were common, the resistance of an electrode T was found with two auxiliary electrodes t1 and t2, a low AC test voltage and an ammeter, measuring each pair in turn. Each pair reads the sum of two electrode resistances, and the three equations solve for T:

Illustrative case. With U = 20 V, i1 = 0.8 A, i2 = 0.5 A and i3 = 0.62 A: the pair resistances are 25.0, 40.0 and 32.26 Ω, and RT = 10 × (1.250 + 1.613 − 2.000) ≈ 8.63 Ω. The method is useful for understanding and for cross-checking, but needs an isolated, current-limited source; never improvise it from the mains. A modern earth tester is safer and filters stray currents.

Two-pole and clamp-on methods, and their limits

Two-pole (dead-earth) method

The tester measures between the electrode and a known low-resistance earth, such as an extensive metallic water main or a large established earth grid. The reading is the sum of both, so it is only meaningful when the reference is far lower than the electrode and well outside its resistance zone. It is quick, needs no long leads, and is acceptable as a screening check on cramped urban sites; it is not an acceptance test.

Clamp-on (stakeless) method

A clamp tester induces a voltage in the conductor it encircles and measures the current that flows round the resulting loop. It therefore reads the electrode under the clamp plus everything else in the loop. It works well where one electrode is connected to a large network of other electrodes in parallel — the rest of the network is then a small addition — and it needs no links opened. Its limits matter on generator sites:

• Single isolated pit: no loop, no reading.
• Two pits linked only to each other (common for DG body earths): the clamp reads the two in series. With pits of 12.4 and 14.0 Ω, it would show about 26.4 Ω — neither pit’s value.
• Pit bonded into a building earth network with, say, 0.8 Ω combined: the clamp reads 12.4 + (14.0 ∥ 0.8) ≈ 13.2 Ω — close to the pit’s true 12.4 Ω, which is where the clamp is useful.
• Metallic loop: if the clamped conductor is part of a loop that closes entirely through metal — for example a pit connected to the earth bar by two conductors, or bonded both by its own conductor and by a buried strip back to the same bar — the clamp shows a fraction of an ohm. That is a continuity result, not an earth resistance.

Use the clamp for routine trending on multi-electrode systems, and the fall-of-potential test for commissioning, acceptance and any doubtful result.

Soil resistivity (Wenner) in brief

Four spikes in a line at equal spacing a; current through the outer pair, voltage across the inner pair. The tester’s reading R gives the apparent resistivity ρ = 2πaR, representative of soil to a depth of roughly a. Keep spike depth small compared with a (about one-twentieth or less). Repeat at increasing spacings, and in two directions at right angles.

Illustrative case. Resistivity is fairly steady to about 4 m and then climbs sharply at 8 m — the signature of rock below a few metres, typical of the granite plateau around Hyderabad. Deep rods would disappoint here; horizontal strips and more shallow electrodes are the better design. Full design use of these figures is on the earth electrode and soil resistivity page and the design calculation guide.

Seasonal correction

Soil moisture changes appreciably to a depth of about 2 m, so shallow electrodes swing the most. In Telangana and Andhra Pradesh the highest readings normally come at the end of the dry season, around April and May, and the lowest during and just after the southwest monsoon. Three practical rules:

• Do acceptance and annual tests in the dry season, on a dry day, without watering beforehand.
• If a test must be done in the wet season, correct it with the site’s own ratio of dry to wet readings from previous years, as in the example above.
• Without history, treat a monsoon reading on shallow electrodes as possibly a third of the dry-season value, and re-test in April before accepting a marginal result.

What the reading means: acceptance thinking

There is no single figure IS 3043 requires for every installation. Specifications commonly ask for about 1 Ω for substations and large installations and a few ohms for DG sets and LV installations; follow the value in your project specification, and remember why it was chosen. The value that matters for shock protection depends on the earthing system: on TN systems the fault loop through protective conductors must be low enough for the breaker to trip — see will your breakers trip on generator supply? — while on TT systems RA × IΔn must not exceed 50 V (25 V on temporary sites). A practical trigger for investigation, whatever the specification: a dry-season reading that has risen by more than about half since the commissioning value, or two pits of the same age and type that differ widely.

Continuity of earthing conductors

An electrode reading means nothing if the conductor from the DG to the pit is broken or loose. Test continuity with a low-resistance ohmmeter that drives a test current of at least 200 mA, zeroing out the leads first:

• Alternator star point to the neutral earth bar, and the neutral bar to each neutral pit (links closed).
• Base frame, canopy, fuel tank, exhaust support and control panel to the body earth bar.
• Body earth bar to each body pit, and to the building main earth bar where the design interconnects them.
• Across every bolted joint you can reach.

For scale, 15 m of 25 mm² copper has a resistance of about 0.727 Ω/km × 0.015 km ≈ 11 mΩ at 20 °C. Readings of a few milliohms to some tens of milliohms on short bonds are normal; a short bond reading more than a fraction of an ohm points to a loose, painted or corroded joint. The bonding layout is set out in earthing, bonding and equipotential architecture.

Exceptions and awkward sites

• No room for 30–50 m leads: use a different direction, run leads along a road or corridor, test before paving, or fall back on clamp or two-pole methods for trending — and say so on the record.
• Large grids and substations: the current spike may need to be hundreds of metres away; these need a planned study, not a routine test.
• Interference: near substations, railway lines or heavy rectifier loads, stray current can disturb readings; test in a second direction at right angles and compare.
• Buried metal nearby (pipes, fences, cable armour) distorts the curve; move the test line away from it.

Earth resistance test record

Common mistakes

• Placing the current spike only 10–15 m away, so the resistance zones overlap and no position of the potential spike gives a true reading.
• Skipping the three-reading check and recording a single 62 % reading as fact.
• Using a clamp meter on two DG body pits linked only to each other and recording their series value as the pit resistance.
• Testing in the monsoon, or after watering, and accepting a marginal result without a dry-season re-test.

What to send us

Send the generator rating and site location, a sketch of the earth pits and what each serves, how much open ground there is around them for test leads, any previous test records with the months they were taken, the earthing specification for the project if there is one, and any symptoms such as earth-fault trips or voltage between neutral and earth. With that we can plan the test, the shutdown needed and the record you will receive. Use the service and AMC enquiry, or WhatsApp the details and photographs.

Related pages

• Earthing & Neutral Grounding
• Earth Pit Types & Maintenance
• Earthing Design Calculation
• DG Set Earthing, Step by Step
• Earth Electrode & Soil Resistivity
• Breaker Tripping on Generator Supply
• Earthing, Bonding & Equipotential Architecture
• Generator AMC

Technical content reviewed by D. Srinivas Reddy, Managing Partner and service engineer. Last updated 3 November 2026.