LV cable sizing: the factor picks the section, not the load

A 160 A feeder needs a 160 A cable. That sentence is wrong in a way that survives into built work, because the current-carrying capacity printed in the table belongs to a cable at 30 °C in still air, alone, and your cable is on a tray with two others in a switchroom at 45 °C. The load did not change. The section did — by two sizes. This guide walks the LV cable sizing tool end to end: what the tables actually promise, how the correction factor is assembled, and the places the selection goes wrong. The worked example is a real run — every figure below was taken from the calculator, not typed in by hand.

What the tool is for

Low-voltage circuits inside the scope of IEC 60364-1, 11.2 a) — nominal voltage up to and including 1000 V a.c. or 1500 V d.c. — sized from the IEC 60364-5-52:2009 Annex B reference ratings.

You give it the design current, or a load in kW with a voltage and a power factor, then the conductor material, the insulation, the number of loaded conductors, one of the ten reference methods A1 to G, the ambient temperature, the group size and arrangement, and — for buried runs — the soil thermal resistivity. It returns the tabulated current the cable must have, the section that first meets it, and a verdict on the section you intend to install.

Above 1 kV this is the wrong instrument. A 1 kV to 36 kV cable is rated from IEC 60502 tables or an IEC 60287 calculation, and once a trench holds more than one row of circuits no tabulated factor describes it at all — that is the subject of the companion guide on ampacity in a real trench.

How the number is produced

One inequality, applied in two directions:

Iz_table  >=  Ib / C          C = Ca x Cg x Cs

Ca  ambient temperature       Table B.52.14 (air, base 30 °C)
                              Table B.52.15 (ground, base 20 °C)
Cg  grouping                  Table B.52.17 (bunched, methods A to C)
                              Table B.52.18 (buried direct, D2)
                              Table B.52.19 (buried in ducts, D1)
                              Table B.52.20 (multi-core in free air, E)
                              Table B.52.21 (single-core in free air, F)
Cs  soil thermal resistivity  Table B.52.16, base 2,5 K.m/W

Forwards, it divides the design current by C and reads down the Annex B column until a row meets it — that gives the smallest section that works. Backwards, it takes the section you actually intend to install, multiplies its tabulated rating by C, and compares the result with Ib. The second number is the one that decides PASS or FAIL, because the cable that goes into the tray is the cable that has to carry the load.

Those two directions produce different-looking numbers from the same run, and confusing them is the first trap. Ib / C is a table current: a figure to look up with, referred to 30 °C still air. It is not a rating, it is not what the cable will carry, and it should never reach a schedule as a capacity.

Every ampacity row and every correction factor in the tool is read from the IEC Tables sub-base of the shared ieccalc database, not hard-coded into the page. Same figures, one place, used by every calculator that needs them.

A worked feeder: 160 A on a tray at 45 °C

A three-phase 400 V feeder, 160 A design current, copper conductors, XLPE/EPR 90 °C insulation, three loaded conductors, multi-core cable on a horizontal perforated tray with two other cables touching it, switchroom ambient 45 °C. Reference method E.

Read the arithmetic rather than the verdict. The tabulated 70 mm² row says 246 A; in this installation that cable carries 175 A, which clears the 160 A load by under 10 %. The 50 mm² row below it says 192 A — comfortably above 160 A if you read the table as a rating, and 137 A once corrected, which is a 23 A shortfall on a cable that a hurried selection would have accepted.

Two conditions come attached to that Cg and are easy to lose. The figure in Table B.52.20 dimensions the arrangement it describes: at least 20 mm from the wall and at least 300 mm between trays. And Table B.52.1 requires, for methods E and F, a clearance to the wall of not less than 0,3 times the cable diameter. Install the tray closer than that and the factor you applied is not the factor for your installation. The tool draws the dimensioned section for the selected method beside the result for exactly this reason.

The same cable, buried

Change nothing about the load and lay the same feeder direct in the ground instead — 45 °C ground temperature, three circuits touching, soil at the table's own 2,5 K·m/W:

Same current, same cable type, and soil exactly as the table assumes — yet the selection moves from 70 mm² to 185 mm². Nothing exotic did that. Cs is precisely 1,000 because 2,5 K·m/W is the base of Table B.52.16; the whole factor is ambient and grouping. And a 95 mm² cable that passed on the tray now fails by 43 A, which is what it looks like when a section is carried across from one part of a route to another.

Where the route changes, and which part governs

That is not a hypothetical. Clause 523.8 of IEC 60364-5-52 puts it plainly:

Where the heat dissipation differs in one part of a route to another, the current-carrying capacity shall be determined so as to be appropriate for the part of the route having the most adverse conditions.

with a NOTE that the requirement "can normally be neglected if heat dissipation only differs where the wiring is going through a wall of less than 0,35 m."

So the tool takes the route as sections, each with its own method, ambient, group and soil, and evaluates every one. A real three-section run for the same 160 A feeder:

The governing section is number 3, and the minimum section for the whole route is 150 mm² — not the 70 mm² that the switchroom, where the cable is visible and where people tend to look, would have justified. Section 2 is marked as a wall pass-through shorter than 0,35 m, so it is excluded from the governing check under the NOTE to 523.8, and the output says so rather than silently dropping it. Installed 95 mm² gives 130,74 A on the worst section: FAIL, by a margin no ambient allowance recovers.

Three more places it goes wrong

The grouping table for a tray is not the grouping table for a bunch. Methods E and F have their own reduction factors, in Tables B.52.20 and B.52.21, and their axis is the support: number of trays or ladders, and number of cables per tray. Table B.52.17 — the familiar 0,80 / 0,70 / 0,65 / 0,60 list for circuits bunched in air, on a surface, embedded or enclosed — does not describe a tray, and applying it there is a different number for a different installation. The tool switches the axis with the method and names the table it read in the result.

Method G has no tabulated grouping factor at all. Table B.52.1 prints a dash for it: single-core cables spaced at least one diameter apart are, in the table's terms, no longer a group. The tool says that explicitly instead of quietly substituting 1,00 — the distinction matters when someone later asks which factor was applied and why.

The installed section has to be a real size. Selections are made on the IEC 60228 preferred series — 1,5 · 2,5 · 4 · 6 · 10 · 16 · 25 · 35 · 50 · 70 · 95 · 120 · 150 · 185 · 240 · 300 mm². A value between two of them has no table row, therefore no rating, and the result will say there is no row rather than interpolate one into existence.

Harmonics: a factor you have to reach for

Third-harmonic current in a four- or five-core cable is handled by Annex E of the same standard, and it is not automatic here — you apply it through the manual C override, because whether the selection is based on the line current or the neutral current is an engineering decision about the load, not something a form field can infer.

Table E.52.1 gives the factors: 0,86 for a third-harmonic content of 15 % to 33 % with the size selected on the line current; 0,86 for 33 % to 45 % with the size selected on the neutral current; and 1,0 above 45 %, still on the neutral current. The consequences are stark in the standard's own example: a 39 A circuit takes 6 mm² clean, 10 mm² at 20 % third harmonic, and 16 mm² at 50 % — where the factor has returned to 1,0 and it is the neutral current, 39 × 0,5 × 3 = 58,5 A, that sizes the cable.

How the implementation is checked

Twelve cases run in the page every time it loads, and each one is a value published in IEC 60364-5-52 rather than a self-check of our own arithmetic:

• Annex E.52.2, the standard's own worked example — 39 A, four-core PVC copper clipped to a wall, method C, selects 6 mm²; the referenced 6 mm² row reads 41 A; with the 0,86 harmonic factor, 39 / 0,86 = 45,35 A moves the selection to 10 mm²
• Correction factors — B.52.14 PVC in air at 40 °C = 0,87 · B.52.15 XLPE in ground at 20 °C = 1,00, the base · B.52.16 direct buried in 1,0 K·m/W soil = 1,50 · B.52.17 five circuits bunched = 0,60 · B.52.20 three cables per tray on two perforated trays touching = 0,80 · B.52.21 two circuits per ladder on two ladders touching = 0,93
• Ampacity rows across the Annex B free-air tables — B.52.10 25 mm² PVC copper in trefoil, method F = 110 A · B.52.12 630 mm² XLPE copper flat and spaced horizontally, method G = 1454 A · B.52.13 300 mm² XLPE aluminium multi-core, method E = 471 A

All twelve pass in the live page. Every figure quoted in the worked examples above was also checked back against the printed table in the standard before it was written down: Table B.52.12 method E (192 / 246 / 298 / 346 / 399 / 456 A for 50 to 185 mm²), B.52.14 (0,87 at 45 °C), B.52.15 (0,89 at 35 °C and 0,80 at 45 °C), B.52.16, B.52.18 (0,65) and B.52.20 (0,82).

What it does not do

Current-carrying capacity is one of several checks a feeder has to pass, and this tool performs that one. It does not size the cable for:

• overload protection coordination — Ib ≤ In ≤ Iz and the operating characteristic of the device, IEC 60364-4-43
• voltage drop — which frequently governs a long LV feeder well before heating does, and which calculator #001 does properly, end to end
• short-circuit withstand — the adiabatic k²S² ≥ I²t check against the prospective fault current at the point of installation
• protective conductor sizing, mechanical protection, fire performance, or any national rule that applies to your installation

Nor does it pretend to rate an installation the Annex B tables do not describe. The reference methods are a closed list of pictures. A duct bank, a mixed trench, a cable on a hot surface, a run through thermal insulation deeper than the table assumes — all of those are outside the list, and the honest answer there is an IEC 60287 calculation on the actual geometry rather than a factor applied to a picture that does not match.

Where the table is the method, and where it stops

For an LV circuit in one of the listed reference installations, the Annex B tables are the design basis. They are exact for the installation they describe, and applying them properly with the right correction factors is not a shortcut — it is the method the standard specifies.

The failure mode is not the table. It is reading a rating as though the conditions in its column heading were not there: 30 °C in air, 20 °C in ground, 2,5 K·m/W soil, 0,7 m depth, one circuit, a stated clearance. Every one of those is a condition your site either meets or does not, and the correction factor is how you find out which.

Run your own feeder →