Tradefox

Tradefox - Renforcer les compétences

Apprendre la construction

BS 7671 Guide · Section 5

Choosing and sizing cables

A cable has to carry its load without cooking, survive a fault long enough for the device to clear it, and deliver usable voltage at the far end. Here’s the design order, what Amendment 4 changed, and the sums to check on every circuit.

The design sequence

Cable sizing is a chain. Do it in order and write the numbers down.

  1. Find Ib, the current the circuit will really carry.
  2. Choose In. Pick a protective device rated at or above Ib.
  3. Find It. Divide In by every correction factor that applies.
  4. Pick the cable whose tabulated rating for your installation method is at least It.
  5. Check voltage drop. Long runs often fail here.
  6. Check fault protection. Confirm Zs gives disconnection in time, and that the CPC survives the fault energy.

Chapter 43 explains why the chain matters. The device must protect the cable from overload, which is the job of Regulation 433.1.1. It must also clear a short circuit before the conductor insulation is damaged, which Section 434 covers. A4 amends Regulation 434.2.1 to use a newly defined term: equipment that is inherently short-circuit proof and earth fault proof. It matters where a smaller conductor is tapped off without its own device, and it’s a design decision, not a site shortcut.

Example: a 9.5 kW shower at 230 V draws about 41.3 A. A 45 A device sits above that, so the cable must be good for at least 45 A once you’ve corrected for its route.

Regulation references: 433.1.1, 434.2.1, 434.5.2, Appendix 4

Current-carrying capacity and installation methods

A cable’s rating depends on how well it loses heat. Appendix 4 handles this with installation methods, each mapped to a reference method with its own column of ratings.

Match the real route to an installation method. If a run passes through several conditions, size for the worst.

What A4 changed

  • Two new installation methods, 72 and 73, have been added.
  • A new reference method, D2, covers cables buried direct in the ground, so direct-buried and ducted cables (D1, following the IEC scheme) are now rated separately.
  • The new methods are referenced in Table 4A2 and in the current-carrying capacity tables for several cable types, including Tables 4D4A, 4E4A, 4H4A and 4J4A.
Watch out: manufacturers’ ratings for buried cables often use different assumptions about soil and depth from BS 7671. Don’t mix the two sets of figures in one calculation.

Regulation references: 523.1, Appendix 4 (Table 4A2 and current-carrying capacity tables)

Correction factors

Tabulated ratings assume 30 °C air (20 °C ground), no neighbouring cables and no insulation. Each correction factor scales the rating to match the real route.

FactorWhat it accounts for
CaAmbient temperature above or below the reference value
CgGrouping with other loaded circuits, which share and trap heat
CiThermal insulation in contact with the cable
CcCircuit or device conditions: a BS 3036 rewireable fuse, or a cable buried in the ground

Why BS 3036 fuses get their own factor

Regulation 433.1.1 requires the current that reliably operates the device (I2) to be no more than 1.45 times Iz. MCBs and cartridge fuses meet this. A rewireable fuse can need about twice its rating to blow reliably, so the cable is oversized using 0.725 (1.45 ÷ 2). Older textbooks call it Cf.

Worked example

A 32 A circuit runs through a plant room at 35 °C, bunched with two other circuits. Suppose the tables give Ca = 0.94 and Cg = 0.70. There is no insulation and the device is an MCB, so Ci and Cc are both 1.

It ≥ 32 ÷ (0.94 × 0.70) = 32 ÷ 0.658 = 48.6 A. You need a cable with a tabulated rating of at least 48.6 A for that installation method, not 32 A.

Same circuit with a 20 A rewireable fuse and no other factors: 20 ÷ 0.725 = 27.6 A tabulated. The weak fuse costs you a cable size.

On site: take the factor values from the current Appendix 4 tables for your exact arrangement and cable type. The figures above only illustrate the method.

Regulation references: 433.1.1, 523.4, 523.5, 523.9, Appendix 4

Thermal insulation

Insulation stops heat escaping, from cables too. Lofts, insulated stud walls and retrofit wall insulation are where cables most often run hot unnoticed.

  • If the cable just touches insulation on one side, use the installation method written for that case. Don’t apply an extra factor on top.
  • If a cable is totally surrounded for more than 0.5 m, Regulation 523.9 lets you take its rating as half the clipped-direct value (Ci = 0.5), unless you have better data.
  • For shorter lengths, such as a cable passing through an insulated stud wall, use the reduced factors in Regulation 523.9.
Watch out: insulation is often added after you’ve left. If the client plans loft or wall insulation, size for it now and note it on the certificate.

Regulation references: 523.9, Appendix 4

Voltage drop

Regulation 525.1 asks that the voltage at equipment is suitable for its standard. Appendix 4 gives the usual way to show this. For an installation fed directly from a public low-voltage supply, keep the drop from the origin to the load within 3% for lighting et 5% for other uses. On 230 V that is 6.9 V and 11.5 V.

Private supplies allow 6% and 8% overall, but final circuits should still meet the public-supply limits. Appendix 4 also allows a little extra on runs over 100 m and during motor starting.

Worked example

A 25 m radial in flat twin and earth carries 28 A. The tabulated drop for 2.5 mm² is 18 mV/A/m.

(18 × 28 × 25) ÷ 1000 = 12.6 V, or 5.5% of 230 V. That fails before you’ve even counted the submain.

Try 4 mm² at 11 mV/A/m: (11 × 28 × 25) ÷ 1000 = 7.7 V, or 3.3%. That leaves 1.7% for everything upstream of the board.

On site: the 5% is a budget for the whole path from the origin. Work out the drop on the meter tails and submain first, then give each final circuit what’s left.

Regulation references: 525.1, 525.202, Appendix 4 section 6

Sizing the CPC: the adiabatic check

During an earth fault, the CPC heats up faster than it can cool. The adiabatic equation in Regulation 543.1.3 finds the smallest CPC that can absorb that energy without damage: S = √(I²t) ÷ k.

  • I is the fault current, roughly U0 ÷ Zs.
  • t is the device’s disconnection time at that current, from its time–current curve.
  • k reflects the conductor material and its insulation. It sets how hot the CPC can get before the insulation suffers. Copper with 70 °C thermoplastic insulation has a higher k than aluminium, and a bare conductor has a different value from one inside a cable. Find the right k in Tables 54.2 to 54.6.

Worked example

Zs is 0.8 Ω on 230 V, so I ≈ 287.5 A. Suppose the device clears in 0.1 s, and k = 115 for a copper CPC in a thermoplastic cable.

S = √(287.5² × 0.1) ÷ 115 = 90.9 ÷ 115 = 0.79 mm². A 1.5 mm² CPC is comfortably adequate.

For very fast devices, use the manufacturer’s let-through energy (I²t) instead of reading a time off the curve. If you’d rather not calculate, a CPC sized from Table 54.7 is accepted without the adiabatic check.

Regulation references: 543.1.3, 543.1.4, Tables 54.2 to 54.7

Cables concealed in walls

A4 brings the rules for hidden cables into one place: Regulation 522.6 and a new Table 52.1. The table sorts walls into two types, without metallic parts and with them (metal studs, for example), and two depths, under 50 mm and over 50 mm from the surface.

The prescribed zones are unchanged in principle:

  • within 150 mm of the top of the wall or partition
  • within 150 mm of a corner where two walls meet
  • running straight horizontally or vertically from an accessory, point or switchgear on the wall

In a wall 100 mm thick or less, the zone carries through to the other side if you can tell where the accessory is from that side.

For a wall without metal parts, a cable less than 50 mm deep needs to be in a zone with 30 mA RCD additional protection, or meet one of the protection methods in 522.6.204, such as an earthed metallic covering or earthed conduit. Deeper than 50 mm, no extra impact protection is needed. Walls with metal parts are stricter: expect to need RCD protection or a 522.6.204 method even for deeper cables. Check Table 52.1 for the exact combination.

Cables under floors and above ceilings are a separate case. Keep them at least 50 mm from the top or bottom of joists and battens, or give them protection, as Regulation 522.6.201 requires.

Regulation references: 522.6, Table 52.1, 522.6.201, 522.6.204, 415.1.1

Terminations for fine-stranded conductors

A4 amends Regulation 526.2 so the connection must suit the conductor class, as listed in Table 52.4. Class 1 is solid and class 2 is stranded rigid. Classes 5 and 6 are flexible, with fine and very fine wires.

Regulation 526.9.1 says fine strands must not be split or splayed. Either use a terminal designed for that class, or treat the end first. A crimped ferrule is the usual fix. Stray strands run hot and can bridge terminals.

On site: check the accessory or device instructions for which conductor classes it accepts. Many screw terminals are only rated for class 1 and 2. Carry a ferrule kit for flex tails, PV and battery cabling.

Regulation references: 526.2, 526.9.1, Table 52.4

Mechanical protection and buried cables

Buried cables need earthed armour or a metal sheath that can act as a protective conductor, unless a conduit or duct gives equivalent protection. They must be deep enough to avoid foreseeable digging, and marked with cable covers or marker tape.

A4 adds a note to Regulation 522.8.10 that settles a long-running argument. Underground conduit with an N750 impact rating, to BS EN 50626-1 or BS EN 61386-24, counts as providing that equivalent protection. Remember the derating: a ducted cable usually needs to be larger than the same cable laid direct, so size it with the right D1 or D2 method.

Regulation references: 522.6, 522.8, 522.8.10, Appendix 4

Related: disconnection times, earthing and bonding conductors, et fire-rated wiring. For a design-calculation refresher, see the Tradefox courses.

Lien secret