In short. An earthing design is a short chain of calculations: measure the soil resistivity, choose an electrode type, calculate the resistance of one electrode and of several, check the result against what the protective devices actually need, size the conductors for the fault current, and add a margin for the dry season. Reddy Generators works through this chain when it installs generators and changeover arrangements across Hyderabad, Telangana and Andhra Pradesh, where the same calculation can give a comfortable answer on black cotton soil and an impossible one on bare granite. This page sets out the method with formulas, two contrasting worked examples and a worksheet.
Who this page is for: Consultants, contractors and plant engineers in Hyderabad, Telangana and Andhra Pradesh who need to turn a soil test into a number of electrodes, a target resistance and conductor sizes for a generator or LV installation, and show their working.
Key facts
- Everything starts from soil resistivity (ρ, in Ω·m). Guessing it is the largest source of error in earthing design.
- Simple formulas are good enough for LV design: one rod R ≈ ρ ÷ L, a vertical plate R ≈ 0.8ρ ÷ perimeter, a buried strip R ≈ 2ρ ÷ L.
- Electrodes only add up properly when they are spaced well apart; closer spacing wastes material.
- The target value comes from the protection: on TN systems the metallic fault loop clears the fault; on TT circuits the rule is RA × IΔn ≤ 50 V (25 V on temporary sites).
- Soil near the surface can be about three times more resistive in the dry season than in the wet. Design for April, not August.
This page is the calculation. Typical resistivity ranges and the ways of lowering resistance are on Earth Electrode and Soil Resistivity Engineering; how the electrodes are physically laid out for a generator is on DG Set Earthing, Step by Step.
Step 1: Measure soil resistivity (Wenner method)
Four test spikes are driven in a straight line at equal spacing a. The instrument passes current between the outer pair and measures voltage between the inner pair, giving a resistance reading R. For spikes driven shallow compared with the spacing:
- Repeat at several spacings, for example 1, 2, 4, 6 and 8 m. The reading at spacing a reflects the soil down to a depth of roughly a, so the series shows whether the soil gets better or worse with depth.
- Take two traverses at right angles across the area where the electrodes will go, clear of buried pipes and cables, which distort the reading.
- Record the date and the weather. A reading in the monsoon is not a design value until it has been corrected.
The formulas below assume uniform soil. If the resistivity changes sharply with depth, use the value at the depth where the electrode will actually sit, and for important sites have a two-layer analysis done.
Step 2: Choose the electrode type
| Soil situation | Usually the better choice | Why |
|---|---|---|
| Deep, uniform soil (alluvium, deep clay) | Vertical rods, driven or with backfill | Cheap, reach stable moisture |
| Soil improving with depth (dry topsoil over moist murrum) | Longer or coupled rods, or drilled electrodes | Puts the metal where the low resistivity is |
| Shallow soil over rock | Horizontal strip or ring in a trench, plates | Rods cannot be driven; length is gained sideways |
| Rock with a moist zone at depth | Drilled borehole electrode with conductive fill | Reaches the weathered, wetter layer |
| Expansive black cotton soil | Rods taken below the cracking zone, with backfill | Surface cracks open in summer and break contact with shallow electrodes |
Step 3: Calculate electrode resistance
where L is the rod length or strip length in metres and P the plate perimeter in metres. These approximations are surprisingly good. For a 3 m rod of 17.2 mm diameter, the more exact textbook formula R = ρ ÷ (2πL) × [ln(4L ÷ r) − 1] gives 0.331ρ, against 0.333ρ from the simple rule. Two useful comparisons follow:
- A 600 × 600 mm plate has a perimeter of 2.4 m, so R ≈ 0.8ρ ÷ 2.4 ≈ 0.333ρ, almost exactly the same as a 3 m rod, and far more digging.
- Backfill compound around a rod works by enlarging its effective diameter. If a 3 m rod sits in a 100 mm column of compound much more conductive than the soil, the exact formula with r = 50 mm gives 0.238ρ, about 28 % lower than the bare rod. Useful, but not magic; it does not turn granite into clay.
Spacing and utilisation. The ρ ÷ (nL) result holds only when rods are more than about four times their length apart. At the more practical spacing of about twice the length, each rod’s zone of influence overlaps its neighbours’. For a handful of rods at that spacing, an illustrative utilisation factor of about 0.85 is reasonable: Rn ≈ R1 ÷ (n × 0.85). Closer spacing lowers the factor further, and doubling the number of rods never halves the resistance in practice.
Step 4: Set the target from the protection
The right target is the value that makes the protection work and keeps touch voltages safe, not a number copied from an old drawing.
- TN systems (most generator installations, with a protective conductor run back to the source): an earth fault returns through the protective conductor and the neutral-to-earth link. Disconnection depends on that metallic loop impedance, which is checked as on Breaker Tripping on Generator Supply. The electrode resistance does not decide whether the breaker trips; it fixes the system’s potential to earth and limits voltages from outside faults and lightning.
- TT circuits (exposed metalwork earthed to a local electrode with no protective conductor back to the source, for example a remote pump house or a site board): the fault current passes through the soil, so a residual current device must disconnect it. The condition is RA × IΔn ≤ 50 V, where RA is the highest value the electrode will reach over the year, including its protective conductor.
For example, a 300 mA RCD allows an electrode of up to about 167 Ω at 50 V. The full table, including the reduced 25 V limit used for temporary sites, is on Earthing and RCD Protection for Temporary and Rental Generator Sites.
Specification practice. IS 3043 does not lay down one universal resistance value. Indian project specifications commonly do: around 1 Ω for substations and large installations, a few ohms for LV generator and building earthing, with individual values stated per pit or for the combined system. Treat these as contract requirements to be met, and treat the protection check above as the engineering requirement that must be met in any case.
Step 5: Size the conductors
Protective conductors in cables are usually sized by the simple rule; earthing conductors and strips are better sized by the adiabatic equation.
| Phase conductor size Sph (same material) | Minimum protective conductor |
|---|---|
| Up to 16 mm² | Same as Sph |
| Over 16 up to 35 mm² | 16 mm² |
| Over 35 mm² | Sph ÷ 2 |
| Conductor | k (initial 30 °C) |
|---|---|
| Copper, separate PVC-insulated conductor (final 160 °C) | 143 |
| Copper, separate XLPE/EPR-insulated conductor (final 250 °C) | 176 |
| Copper core in a multicore PVC cable | 115 |
| Copper core in a multicore XLPE cable | 143 |
| Bare copper, normal conditions (final 200 °C) | 159 |
| Bare steel or GI strip, normal conditions (final 200 °C) | 58 |
The adiabatic size is a thermal minimum. On top of it apply mechanical minimums (buried bare copper is rarely used below 25 mm²; buried GI strip below about 100 mm² and 3 mm thickness corrodes through too soon) and a corrosion allowance for GI. Where a protective conductor must also keep loop impedance low for breaker tripping, the loop check may call for a larger size than the adiabatic one.
Step 6: Allow for the season
Moisture is what makes soil conduct. At about 1 m depth, resistance can vary by roughly three to one between wet and dry seasons. Telangana and Andhra Pradesh have a pronounced swing: soils are saturated from the south-west monsoon (June to September) and at their driest in April and May. Practical rules:
- A design based on a monsoon measurement should multiply the measured resistivity by about 3 for shallow electrodes, unless a dry-season measurement is available.
- Deep electrodes vary less; a rod reaching a permanently moist layer may vary far less than three to one.
- Commissioning readings taken in the wet season will look better than the design. Record the date, and re-test in the dry season before accepting the system as meeting a specification.
- Corrosion and ageing raise resistance over the years; leave a margin of 20–30 % on the target for a long-life installation.
Worked example A: black cotton soil
Illustrative case. A 125 kVA set is to be installed at an agro-processing unit in a black cotton soil belt of northern Telangana. The client’s specification asks for the combined generator earthing system to be no more than 5 Ω in the dry season. Wenner readings were taken in August:
| Spacing a | Reading R | Apparent resistivity |
|---|---|---|
| 1 m | 4.6 Ω | 2π × 1 × 4.6 ≈ 29 Ω·m |
| 2 m | 2.5 Ω | 2π × 2 × 2.5 ≈ 31 Ω·m |
| 4 m | 1.4 Ω | 2π × 4 × 1.4 ≈ 35 Ω·m |
| 6 m | 1.05 Ω | 2π × 6 × 1.05 ≈ 40 Ω·m |
- Interpret: resistivity rises slightly with depth, from wet clay over firmer murrum. At rod depth (3–4 m) the wet value is about 35 Ω·m.
- Seasonal correction: measured in the monsoon, so design ρ ≈ 3 × 35 = 105 Ω·m. Black cotton soil also cracks deeply when it dries, so electrodes must reach below the cracking zone.
- One 3 m rod: 105 ÷ 3 ≈ 35 Ω. With a 100 mm column of backfill: 0.238 × 105 ≈ 25 Ω.
- A 600 × 600 mm plate instead: 0.8 × 105 ÷ 2.4 ≈ 35 Ω, no better than a rod and in the cracking zone. Rejected.
- Four backfilled rods at about 6 m spacing: 25 ÷ (4 × 0.85) ≈ 7.3 Ω. Not enough alone.
- Add a buried ring: 60 m of strip around the plinth at about 1.5 m depth, below the cracking zone: 2 × 105 ÷ 60 ≈ 3.5 Ω.
- Combined: rods and ring in parallel give 2.37 Ω at best; because the rods stand on the ring and inside its area they add relatively little, so take about 3–3.5 Ω, inside the 5 Ω target with modest margin for ageing.
- Conductors: rated current 125,000 ÷ (1.732 × 415) ≈ 174 A; design earth-fault current 8 × 174 ≈ 1,391 A for 1 s. Bare copper minimum 1,391 ÷ 159 ≈ 8.7 mm²; GI 1,391 ÷ 58 ≈ 24 mm², doubled for corrosion ≈ 48 mm². Choose 25 × 6 mm GI (150 mm²) for the ring and the neutral and body conductors, or 25 × 3 mm copper. The output cable has 70 mm² copper phase conductors, so its protective conductor must be at least 70 ÷ 2 = 35 mm².
- A TT outgoing circuit: a pump house 200 m away is earthed to its own backfilled rod, about 25 Ω in the dry season. Protected by a 100 mA RCD: 25 × 0.1 ≈ 2.5 V, well below 50 V.
Worked example B: rocky granite site
Illustrative case. A 62.5 kVA set is to back up a water booster station built on a granite sheet-rock outcrop on the edge of Hyderabad. Readings were taken in April, so no seasonal correction is needed:
| Spacing a | Reading R | Apparent resistivity |
|---|---|---|
| 1 m | 300 Ω | 2π × 1 × 300 ≈ 1,885 Ω·m |
| 2 m | 110 Ω | 2π × 2 × 110 ≈ 1,382 Ω·m |
| 4 m | 40 Ω | 2π × 4 × 40 ≈ 1,005 Ω·m |
| 8 m | 12 Ω | 2π × 8 × 12 ≈ 603 Ω·m |
- Interpret: very resistive near the surface (thin dry soil on rock), improving with depth into fractured, weathered rock where some moisture is held. Near-surface ρ is about 1,900 Ω·m; at depth about 600 Ω·m.
- A 3 m rod (if it could be driven at all) in rock of about 1,400 Ω·m: 1,400 ÷ 3 ≈ 467 Ω. Pointless.
- 80 m of strip in a trench around the compound in the surface layer: 2 × 1,900 ÷ 80 ≈ 47.5 Ω.
- A drilled borehole electrode 150 mm diameter, 20 m deep, with a copper conductor and conductive fill into the weathered zone: the simple rule gives 600 ÷ 20 = 30 Ω; the exact formula with r = 75 mm gives about 29 Ω.
- Strip and borehole together: 17.8 Ω in parallel, divided by 0.9 for overlap: about 20 Ω.
- Is that acceptable? A “few ohms” specification is not achievable here without disproportionate cost, so the design must show safety another way. The set feeds the station as TN-S with a protective conductor in every cable, so a fault on the station’s equipment clears through the metallic loop regardless of the 20 Ω. All metalwork in the compound (set, panel, pump frames, pipework, fence near the set) is bonded together so there is little voltage between things a person can touch at once. Socket and outdoor circuits get 30 mA RCDs; an outlying security-cabin board earthed only to the strip would be TT, and with a 300 mA RCD it needs RA ≤ 50 ÷ 0.3 ≈ 167 Ω, met with a wide margin.
- Conductors: rated current ≈ 87 A; 8 × 87 ≈ 696 A for 1 s needs only 4.4 mm² of bare copper, so the mechanical minimum governs: 25 mm² bare copper or 25 × 3 mm copper strip, which also resists corrosion in the rock fill better than GI.
- Record the reasoning on the drawing: the specification value was not met, why, and what provides protection instead. An honest, reasoned departure is safer than a doctored reading.
Exceptions and limits of the method
- Substations and HV earthing need a proper earthing study for step and touch voltages and ground potential rise, not these LV approximations: see GPR, Touch and Step Studies.
- Strongly layered soil can make the uniform-soil formulas wrong by a large factor; measure at more spacings and use a layered model.
- Lightning protection electrodes have their own requirements; bond them to the power earth but design them to their own standard.
- Electronic and IT systems do not need a “clean” separate earth; they need a low-impedance, bonded one. A separate, unbonded earth is a hazard.
Safety
Soil testing uses long test leads across open ground: keep them clear of traffic, and never drive test spikes where buried cables may run. Tests on an installed system need care so that no part of a live installation is left without an earth; follow Earth Resistance Testing. Design calculations should be checked by a competent engineer, and connections made by licensed people.
Earthing design worksheet
| Design step | Value and working |
|---|---|
| Site, set rating, system (TN-S / TT / mixed) | |
| Wenner readings: spacing and R, with date | |
| Apparent ρ at each spacing (2πaR) | |
| Design ρ at electrode depth (after seasonal factor) | |
| Target resistance and where it comes from (protection / specification) | |
| Electrode type chosen and reason | |
| R of one electrode (formula used) | |
| Number of electrodes, spacing, utilisation factor | |
| Strip or ring length and R | |
| Combined R (with overlap factor) | |
| TT circuits: RA × IΔn check | |
| Earth-fault current and clearing time | |
| Adiabatic minimum size (k used) | |
| Size chosen, with mechanical and corrosion allowance | |
| Dry-season re-test date |
Common mistakes
- Designing from a soil resistivity guessed from a table, or from a single monsoon reading with no dry-season correction.
- Placing rods a metre or two apart and expecting the resistance to fall in proportion to the number of rods.
- Treating the earth-pit reading as what makes breakers trip on a TN system, instead of checking the metallic fault loop.
- Sizing GI strip to the bare thermal minimum with no allowance for corrosion over the life of the installation.
What to send us
Send the site location, the generator or transformer rating and system voltage, the earthing system intended (TN-S, TT or mixed), any Wenner or electrode readings you have with the date they were taken, a sketch of the area available for electrodes, the specification value if one is required, and the protective devices on any remote circuits. We can then review the calculation with you. Use the application engineering enquiry, or WhatsApp the details and photographs.
Questions engineers ask
How do I calculate the resistance of an earth pit?
For a single vertical rod, R ≈ ρ ÷ L, where ρ is soil resistivity in Ω·m and L the rod length in metres. For a vertical plate, R ≈ 0.8ρ ÷ perimeter; for buried strip, R ≈ 2ρ ÷ length. These are approximations for uniform soil and are adequate for most LV design.
How is soil resistivity measured?
By the Wenner four-spike method: four spikes in a line at equal spacing a, current through the outer pair, voltage across the inner pair, and ρ = 2π × a × R. Repeat at several spacings and in two directions to see how the soil changes with depth.
How many earth rods do I need to get below a target value?
Divide the single-rod resistance by the target to get a first estimate, then allow for spacing: at about twice the rod length apart, use a utilisation factor of about 0.85. If the answer is more than a handful of rods, a buried ring or deeper electrodes are usually better.
Why does the earth resistance go up in summer?
Soil conducts through the moisture it holds. In Telangana and Andhra Pradesh the soil dries out from the end of the monsoon to April and May, and shallow electrodes can read about three times higher than in the wet season.
What earth resistance is needed for an RCD-protected circuit?
On a TT circuit, the electrode resistance multiplied by the RCD’s rated residual current must not exceed 50 V, or 25 V on temporary sites. For a 30 mA RCD that allows up to about 1,667 Ω (833 Ω at 25 V); for 300 mA about 167 Ω (83 Ω).
Does chemical earthing compound really reduce resistance?
It enlarges the effective diameter of the electrode and improves contact with the soil, typically reducing a rod’s resistance by around a quarter, and it holds moisture. It cannot overcome very high-resistivity rock on its own.
What k value should I use for a GI earth strip?
For bare steel or GI with a final temperature of 200 °C, k is about 58, against about 159 for bare copper. GI strips are then usually sized at about twice the adiabatic minimum to allow for corrosion.
Related pages
- Earthing and Neutral Grounding
- DG Set Earthing, Step by Step
- Earth Resistance Testing
- Earth Pit Types and Maintenance
- Earth Electrode and Soil Resistivity
- Breaker Tripping on Generator Supply
- Earthing System Study Architecture
- GPR, Touch and Step Studies
Technical content reviewed by D. Srinivas Reddy, Managing Partner and service engineer. Last updated 3 November 2026.