Burning smell in a tea-estate labour quarter after eight months
A two-room labour quarter on a tea estate near Dibrugarh was wired eight months ago with surface PVC conduit, 1.5 mm² copper for lighting and 2.5 mm² copper for sockets. The service connection arrives as a 6 mm² aluminium conductor, and at the meter board the electrician twisted it together with the 4 mm² copper circuit wire, wrapped the twist with insulation tape, and pushed it into the conduit box. The occupant now reports a burning smell near the meter board. Separately, the socket outlet in the inner room is scorched and its plug pin is discoloured. With a 100 W lamp switched on, a multimeter reads 228 V between phase and neutral and 6 V between neutral and earth at that socket. The socket earth wire has been looped on to the iron water pipe of a hand pump, and the house has no earth electrode of its own. Rainfall is heavy and humidity stays above 85% for months.
1. Identify the defects, and decide which defect explains which symptom. Do not assume one fault explains everything.
Read the worked answer
Three separate defects are present; two of them produce different symptoms.
1. The aluminium-copper twisted joint (meter board). This is the burning smell. Quantify the neutral side: the 100 W lamp at 228 V draws 0.44 A; the 6 V between neutral and earth implies about 6 / 0.44 ≈ 13.6 Ω somewhere in the neutral return path. Copper cable of that length would show a fraction of an ohm, so the resistance is concentrated at the bimetallic twist (and partly in the water-pipe path). I²R heating at that joint produces the smell and will eventually carbonise the tape and the box.
2. The neutral-earth voltage itself. It also proves the 'earth' is not at earth potential - the hand-pump pipe is a water path, not an electrode, its joints may be taped or non-metallic, and it is carrying current.
3. The scorched socket. This is a separate, local fault: a loose terminal screw or a worn, enlarged contact in the socket. A discoloured plug pin and a scorch mark at the outlet cannot be caused by a joint at the meter board. It points to poor contact and arcing under a socket load, commonly an iron or a heater.
Judgement. Do not 'fix' the smell by replacing the socket, and do not assume the socket is the only problem. Both terminations must be rebuilt and the earthing replaced before the circuit is returned to service. The installation also fails on the earthing arrangement alone, regardless of the burning smell.
2. Explain why the aluminium-to-copper twist fails specifically in this climate, and describe how the joint must be made instead.
Read the worked answer
Why it fails here.
- Galvanic couple. Aluminium and copper sit far apart in the galvanic series (roughly 2 V apart). In the presence of moisture - and at 85% RH for months there is always moisture - the joint becomes a small electrolytic cell, and the aluminium, being anodic, corrodes preferentially and pits.
- Oxide film. Aluminium forms an insulating oxide within minutes of being abraded, so contact resistance is high from the start unless the oxide is broken and kept out.
- Differential expansion. The two metals expand at different rates, so every load cycle works the joint loose; it creeps and the contact pressure falls.
- Environment. Heavy rainfall, high humidity and the warm, slightly corrosive estate atmosphere accelerate all three effects.
The result is a joint whose resistance rises, heats, oxidises further, and finally arcs - a classic fire sequence, not a nuisance fault.
How it must be made.
- Use a bimetallic (copper-aluminium) transition connector or an aluminium-rated compression lug, sized for 6 mm² aluminium and 4 mm² copper.
- Abrade the aluminium, apply an oxide-inhibiting compound, and crimp with the correct die; do not solder aluminium with ordinary solder.
- Torque the terminals to the maker's figure and re-check after the first thermal cycle.
- Enclose the joint in a proper, dry junction box - never a taped twist pushed into a conduit box.
- Better still, terminate the aluminium at the meter with an approved connector and run copper from there.
Twist-and-tape is not a jointing method at all: it has no contact pressure control, no oxide control, and no protection from the couple.
3. List the tests that must be passed before re-energising, and state how the earthing must be corrected.
Read the worked answer
Tests before re-energising.
- Insulation resistance at 500 V DC between phase-neutral, phase-earth and neutral-earth: 1 MΩ or more throughout, including the damaged section.
- Earth continuity: resistance from the socket earth pin back to the electrode, expected well under 1 Ω.
- Earth electrode resistance by the three-pole fall-of-potential method; for a small installation a value of about 8 Ω is the usual working target, and lower is better.
- Polarity: phase to the centre pin of the socket, neutral to the outer, switch in the phase conductor.
- Neutral-earth voltage under load: should be near zero, at most a volt or two. A reading of 6 V must disappear after the neutral joint is rebuilt.
- RCD trip test at 30 mA, both by test button and with a dedicated tester for trip time.
Earthing correction.
- Install a dedicated electrode: a 40 mm GI pipe driven at least 3 m into permanently moist ground, or a copper plate in an earth pit, sited away from the apron and the drain.
- Backfill in the recognised way - alternating layers of charcoal and salt in the pit - and bring up a protected earth conductor of at least 2.5 mm² (4 mm² is better) to the board and to each socket.
- Bond the metal water pipe to the earthing system as supplementary bonding if it is metallic and electrically continuous, but it must never be the sole earth path.
- Test after the soil has settled and again in the dry season, because electrode resistance rises when the ground dries out.