High Earth Resistance at a Small Workshop
A small 415 V workshop feeds a 3.7 kW pump, a lathe and lighting from a distribution board whose earth bar is connected to one 40 mm diameter GI pipe electrode driven 3 m into the ground, with its top 0.5 m below grade. The installation has 32 A MCBs but no RCCB, and all machine frames are bonded to the earth bar. In May, a fall-of-potential test with the electrode disconnected from the bar gives 14.6 Ω with the potential probe at 20 m and 14.9 Ω at 40 m. The local utility insists on 5 Ω or less for a low-voltage installation. The soil is sandy loam that dries out between April and June; bedrock is reported at about 5 m depth. The electrician proposes driving one more pipe and 'watering it well'.
1. Assess the electrician's proposal with numbers: how many electrodes are actually needed, and what spacing and interconnection would you specify?
Read the worked answer
Calculate, don't guess.
- Two electrodes in parallel: 14.6/2 = 7.3 Ω - above 5 Ω.
- Three: 14.6/3 = 4.87 Ω ideal, but electrodes 6 m apart still interfere; apply a practical mutual-coupling factor of about 1.15 → 5.6 Ω - still above 5 Ω.
- Four: 14.6/4 = 3.65 Ω × 1.15 ≈ 4.2 Ω - acceptable, with margin for the dry season.
So three additional pipes, not one. Specify:
- 40 mm GI pipe, 3 m long, spaced 6 m apart (at least twice the driven depth), in a line or L-shape, tops 0.5 m below grade.
- Interconnect with 25 × 6 mm GI strip, joints cleaned and bolted with two bolts or brazed, then coated with bitumen.
- Backfill each with bentonite or a soil–charcoal mixture, not plain sand; water during dry months.
- Re-test with the electrode disconnected; accept when ≤ 5 Ω in the dry season.
If bedrock stops the pipes short of 3 m, use a longer rod in a bored hole or a chemical electrode rather than adding more shallow pipes, which couple strongly and add little.
2. What is wrong with judging the earth by the May reading alone, and how should the test be performed?
Read the worked answer
A May reading is the best news, not the worst: resistance rises as soil dries, so a dry-season value of 14.6 Ω is the worst case. The bigger trap is the opposite - measuring with the electrode still bonded to the earth bar gives a falsely low reading because other paths (water pipe, building steel, neighbouring earths) are in parallel.
Correct method:
- Disconnect the electrode under test from the installation.
- Use a three-pole fall-of-potential tester; drive the current probe at least 5 × the electrode depth away (≥ 15 m; 20 and 40 m as done), and the potential probe at about 62 % of that distance. Repeat with the potential probe moved ±5 m; results should agree within a few percent.
- Test in more than one direction and record soil condition and temperature.
- Test in the dry season and again in the monsoon; log both. The installation is judged by the worst reading.
- Check tester calibration and lead resistances; a loose lead can read either high or low.
3. Until the earth is improved, what is the actual danger, and what interim protection is required?
Read the worked answer
With 14.6 Ω, a phase-to-frame fault is not cleared quickly by the 32 A MCB. Approximate fault current = 230/14.6 ≈ 15.8 A (ignoring the small transformer and conductor impedance). A 32 A MCB needs roughly 5–10 times rated current for magnetic operation, i.e. 160–320 A, so it will not trip instantly. The fault current flows through the earth path and the machine frame rises close to line voltage until the thermal element eventually operates - a shock hazard, not just a 'bad earth'.
Interim measures:
- Install a 30 mA RCCB on the affected circuits; it trips at 30 mA leakage and clears the fault in milliseconds.
- Verify main equipotential bonding of the panel, machine frames and any metallic water pipe.
- Size the earthing conductor to IS 3043 Table 1 (for a 6 mm² phase conductor, at least 6 mm²; the buried electrode lead should be 25 × 6 mm GI strip or 8 SWG GI wire).
- Mark the installation as having earth resistance above limit and restrict use until the electrodes are added.
The RCCB is mitigation, not a substitute: the 5 Ω target must still be met.