Transformer Earthing System: System Earthing, Protective Earthing, Lightning Earthing, the ≤4 Ω Rule, Soil Resistivity Testing & Ground Enhancement

Abstract

The earthing (grounding) system of a substation and its power transformers serves three distinct purposes — system earthing (providing a reference point for the power system and controlling earth-fault currents), protective earthing (limiting touch and step potentials to safe levels for personnel and providing a low-impedance path for fault current to operate protection), and lightning earthing (dispersing lightning surge energy into the earth without developing excessive ground potential rise). The ≤4 Ω ground resistance rule — one of the most frequently cited and misunderstood numbers in electrical engineering — is examined in detail: its origin, its applicability, and the cases where it is insufficient or unnecessarily stringent. This article covers the classification of transformer earthing (solid, resistance, reactance, isolated), the Wenner four-pin soil resistivity measurement method, ground resistance testing methods (fall-of-potential, clamp-on), and ground enhancement techniques (chemical electrodes, bentonite, ground enhancement material, counterpoise).

1. The Three Functions of Transformer Earthing

1.1 System Earthing (Neutral Earthing)

The transformer neutral point is connected to the earth electrode system. The earthing method determines:

• The magnitude of earth-fault current — and thus the sensitivity and selectivity of earth-fault protection
• The temporary overvoltage (TOV) on healthy phases during an earth fault (earth fault factor)
• The insulation level required for the transformer windings

| Earthing Method | Earth Fault Current | Earth Fault Factor | Application | |-----------------|---------------------|--------------------|-| | Solid (direct) | High (I_f = I_L-G, limited by system impedance) | ≤1.4 | HV transmission systems (≥110 kV) | | Low-resistance | Controlled (100-1,000 A) | 1.4-1.73 | MV industrial systems (6.6-33 kV) | | High-resistance | Low (≤10 A, ≤5 A for alarm-only) | Near 1.73 | MV process-critical systems, generator neutrals | | Reactance (Petersen coil) | Near zero (compensated by coil inductance) | Near 1.73 | MV distribution (mostly Europe) | | Isolated (ungrounded) | Capacitive only (small) | 1.73-3.0 | Special applications (limited distribution) |

For a 132/33 kV substation transformer:

• HV neutral (132 kV): Solidly earthed — this is a transmission system, direct earthing is standard per the grid code
• LV neutral (33 kV): Typically resistance-earthed through a neutral earthing resistor (NER) to limit the single-line-to-ground fault current to 300-1,000 A, reducing equipment damage and minimizing arc-flash energy while maintaining sufficient current for protection operation

1.2 Protective Earthing (Equipment Earthing)

All non-current-carrying metal parts of the transformer, switchgear, and supporting structures are connected to the substation earth grid. The purpose is:

• To maintain the potential of all exposed metalwork at the same level under earth-fault conditions (equipotential bonding)
• To provide a low-impedance return path for earth-fault current, ensuring that the protection operates within the specified fault-clearing time
• To limit touch potential (potential difference between a grounded structure that a person can touch and the earth surface at the person's feet) and step potential (potential difference between a person's feet 1 m apart) to safe levels

Safe limits per IEEE 80 / IEC 60479 (for a 50 kg body):

E_touch_allowable = (1,000 + 1.5 × C_s × ρ_s) × 0.116 / √t_f E_step_allowable = (1,000 + 6.0 × C_s × ρ_s) × 0.116 / √t_f

Where:

• C_s = surface layer de-rating factor (typically 0.5-1.0, lower for high-resistivity crushed rock)
• ρ_s = surface layer resistivity (Ω·m) — crushed rock 2,000-5,000 Ω·m is standard in substations
• t_f = fault clearing time (s)

For a typical substation with crushed rock surface (ρ_s = 3,000 Ω·m, C_s = 0.7), and fault clearing time t_f = 0.5 s:

E_touch = (1,000 + 1.5 × 0.7 × 3,000) × 0.116 / √0.5 = 4,150 × 0.164 = 681 V E_step = (1,000 + 6.0 × 0.7 × 3,000) × 0.116 / √0.5 = 13,600 × 0.164 = 2,232 V

The substation earth grid design must ensure that the actual touch and step potentials during an earth fault do not exceed these calculated limits.

1.3 Lightning Earthing

The lightning earth electrode system must:

• Disperse the lightning current (typically 10-100 kA peak, 10/350 μs waveform) into the earth
• Control the ground potential rise (GPR) such that back-flashover does not occur (surge transferred from ground to the phase conductor through the arrester or through insulation breakdown)
• Respond to the high-frequency content of the lightning surge (≤1 MHz) — the inductive impedance of long earth electrodes may dominate over the resistive impedance at lightning frequencies

The effective length (l_eff) of an earth electrode at lightning frequencies:

• For a single vertical rod: l_eff ≈ 10-20 m (the inductive impedance of a deeper rod negates the benefit of additional depth for lightning)
• For a horizontal counterpoise (buried conductor): l_eff ≈ 20-30 m per direction from the injection point

2. The ≤4 Ω Rule: Origin, Applicability, and Limitations

2.1 Origin

The 4 Ω rule originates from historical distribution transformer earthing requirements. For a 240 V single-phase supply (common in North American residential distribution), the neutral is required to have a ground resistance ≤25 Ω per the NEC. For a transformer pole ground, ≤25 Ω at each pole, with multiple poles in parallel typically achieving an overall resistance of ≤4 Ω.

In practice, the 4 Ω number became embedded in utility standards worldwide as a one-size-fits-all requirement. The reality is more nuanced.

2.2 When ≤4 Ω Matters

The ground resistance of a substation mesh is relevant when the earth fault current is supplied solely through the ground return path. For solidly earthed HV systems, the earth fault current returns primarily through the overhead earth wire (OHGW) or cable sheath connected to adjacent substation earth grids — the local ground resistance contributes only a fraction of the total return impedance.

For a 132 kV substation with a 10 kA earth fault current:

• If the station ground resistance is 1 Ω: GPR = 10 kA × 1 Ω = 10 kV
• If the station ground resistance is 0.1 Ω: GPR = 10 kA × 0.1 Ω = 1 kV

The difference in GPR is dramatic. However, the ground resistance is not the sole determinant of safety — the grid design (conductor spacing, depth, surface material, equipotential mesh) may achieve safe touch and step potentials even with a relatively high ground resistance.

2.3 When ≤4 Ω Is Not Sufficient

A 4 Ω ground resistance with a 10 kA earth fault current produces a GPR of 40 kV — which will certainly cause dangerous touch potentials and potential transfer (the earth potential rise is exported through communication cables, LV neutrals, pipelines, and rail tracks to remote locations where the local earth potential is near zero). In this case, 4 Ω is grossly inadequate — the substation needs ≤0.5 Ω.

The correct criterion: The ground resistance must be low enough that the product of ground resistance and earth-fault current (GPR = I_f × R_g) does not exceed:

• The insulation withstand of control, communication, and LV cables leaving the substation (typically 2-5 kV for standard LV cables)
• The allowable touch and step potentials per IEEE 80 / IEC 61936-1

3. Soil Resistivity Measurement

3.1 Wenner Four-Pin Method

The Wenner method (IEEE 81, ASTM G57) is the standard for measuring soil resistivity. Four electrodes are driven into the soil in a straight line at equal spacing (a). A known current I is injected through the outer electrodes (C1, C2), and the voltage V is measured between the inner electrodes (P1, P2).

ρ = 2π × a × (V / I) (for a homogeneous half-space, and a is small compared to the depth of interest)

For a uniform soil, the apparent resistivity at electrode spacing a is an approximation of the average resistivity to a depth of a. By varying the spacing a (typically 0.5, 1, 2, 4, 8, 16, 32 m), a vertical resistivity profile is obtained.

Multi-layer interpretation:

• Shorter spacings (a = 0.5-2 m) → near-surface resistivity (relevant for step potential surface layer)
• Medium spacings (a = 4-8 m) → intermediate depth (relevant for grid conductor burial depth, typically 0.5-1 m)
• Long spacings (a = 16-32 m) → deep soil resistivity (relevant for deep-driven rods and the overall GPR)

The raw apparent resistivity data is processed using a soil-modeling software (e.g., CDEGS, ETAP Ground Grid) to fit a two-layer or multi-layer soil model. The two-layer model provides:

• ρ₁ = upper layer resistivity (Ω·m)
• h₁ = upper layer thickness (m)
• ρ₂ = lower layer resistivity (Ω·m, often assumed infinite depth)

3.2 Typical Soil Resistivity Values

Soil Type	Resistivity (Ω·m)	Ground Resistance Implication
Sea water	0.1-1	Excellent — near-zero resistance achievable
Marsh, wet organic soil	10-50	Very good
Moist clay, loam	50-100	Good
Dry clay, sandy clay	100-500	Moderate — grid design required
Sand and gravel (dry)	500-5,000	Poor — ground enhancement needed
Rock (granite, basalt)	1,000-10,000+	Very poor — specialized design
Permafrost, dry desert sand	10,000-100,000+	Extremely poor — may require counterpoise + chemical treatment

4. Ground Enhancement Methods

4.1 Deep-Driven Rods

Driven vertical copper-clad steel rods (typically 16-20 mm diameter, 3-6 m length) penetrate through the high-resistivity surface layer to reach lower-resistivity deep soil. When Wenner measurements show ρ₂ < ρ₁ (lower resistivity at depth), deep-driven rods are the most effective enhancement.

4.2 Ground Enhancement Material (GEM)

Conductive backfill material — typically a bentonite-based compound mixed with graphite or carbon — is placed around the ground electrode to lower the contact resistance between the electrode and the surrounding soil. Bentonite swells when wet, maintaining a conductive gel in contact with both the electrode and the native soil.

Application: Drill or auger a hole 100-200 mm diameter, insert the copper-clad steel rod, fill the annulus with GEM slurry, and compact. The GEM reduces the electrode-to-soil contact resistance by 40-70% in sandy/gravelly soils.

4.3 Chemical Electrodes

Chemical ground electrodes are hollow copper tubes filled with hygroscopic salts (e.g., magnesium sulfate, sodium chloride) that leach into the surrounding soil through weep holes, lowering the local soil resistivity. Used when:

• The native soil is dry and highly resistive (rock, desert sand)
• The climate is arid and soil moisture is consistently low

Disadvantage: The salt leaches out over 5-10 years, requiring electrode replacement. The salt can also corrode nearby buried metal structures if not properly contained.

4.4 Counterpoise (Horizontal Buried Conductor)

A counterpoise is a buried bare copper conductor (typically 70-120 mm²) extending radially from the substation in 2-4 directions, each 20-50 m long. The counterpoise provides a large surface-area contact with the soil, distributing the current and lowering the overall resistance. Counterpoise is also effective at lightning frequencies because the distributed inductance-capacitance network behaves as a transmission line, more efficiently dispersing the high-frequency energy than a single concentrated electrode.

FAQ

Q: Where does the ≤4 Ω rule come from, and is it still valid?

The ≤4 Ω rule originated in the early 20th century for distribution transformer pole grounds. For a single-phase 240 V system with a neutral grounded through a 4 Ω electrode, the maximum ground-fault current is 240 / 4 = 60 A — insufficient to blow a typical service fuse quickly, but sufficient to be detected. The rule persists because it is simple, easily tested, and provides a bright-line compliance test. For modern HV substations, the ≤4 Ω rule is not a sufficient safety criterion by itself — touch and step potential calculations per IEEE 80 / IEC 61936-1 must be performed, and the ground resistance is only one input. A substation with R_g = 0.1 Ω may still have dangerous touch potentials if the grid design is poor (too-coarse mesh, no surface gravel, no equipotential bonding of the fence). Conversely, a substation with R_g = 5 Ω may be safe if the mesh is fine, the surface material is high-resistivity crushed rock, and the fault clearing time is fast.

Q: How do I test substation ground resistance?

The Fall of Potential method (IEEE 81) is the standard: (1) disconnect the substation ground from the overhead earth wire (OHGW) and transformer neutrals (isolate the grid), (2) drive a current probe (C2) 5-10× the maximum diagonal dimension of the grid away from the grid perimeter, (3) drive a potential probe (P2) between the grid and C2, at the 61.8% point (the 62% rule — at approximately 0.618 × distance from grid to C2, the measured potential is the true grid potential rise), (4) inject a test current (~10-50 A, typically at 70-128 Hz to avoid 50/60 Hz interference) and measure the potential difference between the grid and P2. The ground resistance R_g = V / I. For large grids (>100 m diagonal), the current probe must be several hundred meters away, which may not be physically achievable — in that case, the Slope method or computer modeling is used to derive R_g from the fall-of-potential curve.

Q: What is the difference between earth resistance and earth resistivity?

Earth resistivity (ρ, unit: Ω·m) is a material property of the soil — it describes how much the soil resists the flow of electric current per unit volume. It depends on soil type, moisture content, temperature, and salt concentration. Earth resistance (R, unit: Ω) is the resistance of a specific electrode configuration to remote earth — it depends on the soil resistivity AND the electrode geometry (length, diameter, depth, number of parallel electrodes). For a single vertical rod of length L and diameter d, buried in homogeneous soil of resistivity ρ: R ≈ (ρ / 2πL) × ln(4L/d). A longer rod in the same soil has lower resistance — this is why you test soil resistivity first, then design the electrode geometry to achieve the target resistance. You cannot "measure ground resistance" without an installed electrode system — the fall-of-potential test measures the resistance of an existing grid.

Q: Why is the transformer neutral earthed separately from the substation grid?

It is not — the transformer neutral is connected to the substation earth grid. "Separately derived system" refers to the fact that the transformer secondary neutral is a new system neutral (for the LV system) distinct from the primary system neutral, and each must be earthed per the applicable code. Both the HV and LV neutrals connect to the same substation earth grid. In some designs, the neutral earthing resistor (NER) is located in a separate enclosure with its own connection to the grid, but this connection ultimately ties to the same grid.

Q: Should I bond the substation fence to the earth grid?

Yes — absolutely. A metallic fence within the grid area or on the grid perimeter must be bonded to the earth grid at every fence post or at least every 3-5 meters. During an earth fault, the grid potential rises (GPR). If the fence is not bonded, the potential difference between the fence and the surrounding earth can reach thousands of volts — a lethal touch potential for anyone touching the fence. If the fence is bonded, the fence rises to the same GPR as the grid, and the touch potential is between the fence (at GPR) and the earth surface (at some fraction of GPR), which is controlled by the grid design. For fences on the grid perimeter, a separate perimeter conductor should be buried 0.5-1 m outside the fence, bonded to the fence and the grid, to grade the potential (potential control ring).

Q: Can I use the transformer tank itself as the earth electrode?

No. The transformer tank is an above-ground metallic enclosure resting on a concrete plinth or steel support. Its contact with the earth is unreliable (painted surface, air gap at the foundation interface) and it does not constitute an earth electrode. The transformer tank, core, and all metalwork must be connected to the earthing system by dedicated copper earthing conductors of adequate cross-section (typically ≥70 mm² Cu for lightning current-carrying connections, ≥35 mm² for protective earth bonding). The earthing conductor from the transformer tank to the grid is tested during commissioning — typically 25 A DC is injected and the voltage drop is measured to verify the connection integrity.

References / Standards

Reference	Title
IEEE 80-2013	IEEE Guide for Safety in AC Substation Grounding
IEEE 81-2012	IEEE Guide for Measuring Earth Resistivity, Ground Impedance, and Earth Surface Potentials of a Grounding System
IEC 61936-1:2014	Power installations exceeding 1 kV a.c. — Part 1: Common rules
IEC 60479-1:2018	Effects of current on human beings and livestock — Part 1: General aspects
IEC 60076-1:2011	Power transformers — Part 1: General
NFPA 70 (NEC 2023)	Article 250 — Grounding and Bonding
ASTM G57-20	Standard Test Method for Measurement of Soil Resistivity Using the Wenner Four-Electrode Method

*Authored by Du Fu, Production Engineer at ZY POWER. Earthing system design is a safety-critical discipline — always verify the design with a formal touch-and-step potential analysis per IEEE 80 and test the installed ground resistance per IEEE 81 before energizing the transformer.*