Installation practice, testing and verification
Installation practice
Selection and installation
Sound installation practice starts with the load: calculate maximum demand, then select a cable whose current-carrying capacity, after every applicable derating factor, is at least the rating of the protective device. AS/NZS 3008 gives ratings for each installation method (clipped direct, enclosed in a wall, in conduit, buried) plus correction factors for ambient temperature (C_a), grouping (C_g), thermal insulation (C_i) and soil conditions. If the derated capacity falls below the device rating, increase the cable size, change the installation method or reduce the load.
Mechanical protection, support, bending radius, identification and accessibility all apply. Cables must not be damaged by the building fabric, must be supported at the required intervals, must be identified at terminations and junctions, and equipment must remain accessible for maintenance. Enclosures need an IP rating suited to the environment, for example finger-safe internally and weatherproof outdoors. Single-phase identification is brown active, blue neutral and green/yellow earth; three-phase is brown, black and grey. Switchboards require a main switch, MEN link, RCDs and clear circuit identification.
Traps
Forgetting thermal insulation (C_i) - a cable buried under ceiling insulation can lose close to half its rating. Using a free-air table for a cable in a wall cavity. Treating the protective earth as a current-carrying conductor when it is not. Choosing cable by nominal size rather than by derated capacity. Mixing old and new colour codes at a junction.
Worked example
A 2.5 mm² TPS cable clipped direct is rated about 20 A. At 40 °C ambient (C_a ≈ 0.87) grouped with two other circuits (C_g ≈ 0.70), capacity = 20 × 0.87 × 0.70 ≈ 12.2 A, so a 16 A breaker would not be acceptable for that run.
Installation practice, testing and verification
The verification process
AS/NZS 3000 Clause 8 and AS/NZS 3017 set out the process. Verification begins with visual inspection - everything is confirmed correct before any instrument is used. Then, with the installation isolated and proved dead, tests are done in a set order: continuity of the protective earthing conductor and equipotential bonding (commonly 0.5 Ω or less); insulation resistance at 500 V DC (minimum 1 MΩ); polarity; correct circuit connections; then, once re-energised, fault loop impedance and prospective short-circuit current, and RCD operation. Functional testing of switches, controls and equipment follows. Results are recorded and the work is certified.
Instruments
Prove the tester before and after testing, using a known live source or a proving unit. A tester that has not been proved tells you nothing about the circuit.
Traps
Insulation resistance testing with sensitive electronics, surge diverters or dimmers still connected gives a false fail and may damage equipment. The 1 MΩ criterion is a minimum, not a target. Fault loop results must be compared with the maximum permitted value for the actual protective device, not a generic figure. Every RCD must be tested individually, including its test button and an instrument test. And no voltage is not the same as isolated - the circuit must be proved dead at the point of work.
Worked example
A 20 A Type C MCB protects a 230 V final subcircuit. A Type C device trips magnetically at up to 10 × I_n = 200 A, so the maximum earth fault loop impedance is Z_s = V/I_a = 230/200 = 1.15 Ω. A measured value above that means the circuit will not disconnect in time; a Type B device (5 × I_n) would allow 2.3 Ω.