Derated current-carrying capacity to BS 7671 and IEC 60364 — Iz = It × Ca × Cg × Ci × Cf, with the ambient, grouping, thermal insulation and BS 3036 factors applied and the Ib ≤ In ≤ Iz condition checked.
A cable’s tabulated current-carrying capacity It assumes 30 °C ambient, one circuit on its own, no thermal insulation and an ordinary protective device. The capacity you may actually use is Iz = It × Ca × Cg × Ci × Cf, where Ca corrects for ambient temperature, Cg for grouping, Ci for thermal insulation and Cf for a BS 3036 semi-enclosed fuse. Working backwards, the tabulated capacity you must buy is It ≥ In ÷ (Ca × Cg × Ci × Cf). Three circuits bunched in a 35 °C loft give Cg = 0.70 and Ca = 0.94, a combined factor of 0.65 — so a 32 A device needs a cable tabulated at 32 ÷ 0.65 = 49 A, not 32 A. The whole design must satisfy Ib ≤ In ≤ Iz.
The factors are the same — BS 7671 Tables 4B1 and 4C1 derive from IEC 60364-5-52. Only the naming changes. For NEC ampacity derating, use the wire size calculator.
The free-air, 30 °C figure for your cable and reference method, from BS 7671 Appendix 4 or the manufacturer’s data. It depends on the installation method as much as the conductor, which is why it is not a dropdown here.
The current the circuit will actually carry.
The nominal rating of the protective device. Must be at least Ib and no more than Iz.
Sets the conductor operating temperature θc used in Ca.
The air temperature around the cable in service, not today. 30 °C is the tabulated reference.
The standard permits Ca above 1.00 in cool conditions. Taking it makes the design depend on an ambient nobody can guarantee for the life of the installation.
Only the bunched-and-enclosed column is built in. Single-layer arrangements on wall, ceiling, tray or ladder are in the same table but could not be verified to this site’s standard — enter those yourself.
Regulation 523.9. The 0.5 factor is the one that arrives later, when the loft gets topped up over a cable that was fine when it was installed.
A semi-enclosed rewireable fuse requires It ≥ In ÷ 0.725.
Ib ≤ In ≤ Iz is satisfied: 20 A ≤ 20 A ≤ 27.0 A. The cable has 7.0 A of headroom over the design current.
Current capacity is only half of cable selection. Run the same circuit through the voltage drop check and take whichever answer gives the larger conductor — on runs over about 25 m, voltage drop usually decides.
The everyday case, and the one that catches people out. Three circuits in trunking through a roof space that reaches 35 °C in summer.
Ca 0.935 × Cg 0.70 = 0.655. Iz = 17.7 A against a 32 A device — it fails. Needs It ≥ 48.9 A.
A cable on its own at 30 °C with nothing else going on. Every factor is 1.00 and Iz is simply It — the case the tables are written for and the case you almost never have.
Combined factor 1.000. Iz = 27 A, 20 A ≤ 20 A ≤ 27 A — passes with 7 A spare.
A 2.5 mm² ring leg that was clipped to a joist when the house was built, and disappeared under 300 mm of mineral wool twenty years later. Regulation 523.9 halves it.
Ci 0.5. Iz = 13.5 A against a 32 A device — badly protected.
A BS 3036 semi-enclosed fuse lets an overload run far longer before clearing, so the cable has to be bigger for the same nominal rating.
Cf 0.725. Iz = 19.6 A against a 30 A fuse — needs It ≥ 41.4 A.
Where 90 °C thermosetting insulation earns its money: at 45 °C ambient its Ca is 0.87 against 0.79 for thermoplastic.
Ca 0.866 × Cg 0.80 = 0.693. Iz = 58.9 A — just short of the 63 A device.
When to use this: before you buy the cable, whenever the circuit is going to share a conduit or trunking with others, whenever the run passes through a loft, a plant room or an airing cupboard, or whenever you are working on an older board with rewireable fuses.
Every current-carrying capacity in a cable table is quoted for a set of conditions that a real installation almost never meets: 30 °C ambient, the circuit on its own, no thermal insulation anywhere near it, and a protective device that operates decisively. Change any of those and the cable runs hotter for the same current. The regulations handle this with four multipliers:
Iz = It × Ca × Cg × Ci × Cf
It is the tabulated capacity, Iz is what you may actually use, and the four C terms are ambient temperature, grouping, thermal insulation and protective device type. When you are choosing a cable rather than checking one, the same relationship runs backwards:
It ≥ In ÷ (Ca × Cg × Ci × Cf)
The factors compound, and that is what surprises people. Three circuits bunched in a 35 °C loft is not a 30% problem and a 6% problem; it is 0.70 × 0.94 = 0.65, so a cable tabulated at 27 A may carry 17.7 A. A 32 A circuit in those conditions needs a cable rated 49 A in free air — two or three sizes up from what the breaker suggests.
Three numbers in order, and both inequalities do real work.
In ≥ Ib keeps the device from operating in normal service. In ≤ Iz keeps the cable from being the thing that fails first. Get the second one wrong and there is no alarm: the circuit works, the breaker never trips, and a sustained overload quietly ages the insulation for years.
A cable sheds heat into the air around it, so the hotter that air, the less current the cable can carry before its conductor reaches the temperature its insulation is rated for. The factor is:
Ca = √[(θc − θa) ÷ (θc − 30)]
θc is the conductor’s permitted operating temperature — 70 °C for thermoplastic insulation, 90 °C for thermosetting — and θa is the ambient. The square root is not a fudge: heating in a conductor goes with I²R, so the current that produces a given temperature rise goes with the square root of the headroom you have left.
The temperature that matters is the one the cable will see in service, not the one on the day you install it. A domestic loft in the UK regularly passes 40 °C in summer; a boiler cupboard or a plant room can sit above that all year. Designing on a comfortable spring afternoon is how a compliant circuit becomes a non-compliant one in July.
The formula returns a factor above 1.00 for ambients under 30 °C, and the standard permits you to take it. This calculator holds Ca at 1.00 unless you explicitly ask for the uplift, because ambient temperature is the input a designer is least able to guarantee for the life of an installation. Refusing the credit can only ever produce a slightly larger cable. Taking it means the design is only correct while the room stays cool, and rooms change use.
Cables next to each other heat each other. Each one is trying to dump heat into the same air, and the ones in the middle of a bunch have nowhere to send it. The factor for cables bunched or enclosed together — in conduit, in trunking, in ducting, or simply clipped together on a surface — runs:
| Circuits | 1 | 2 | 3 | 4 | 5 | 6 | 9 | 12 | 20 |
|---|---|---|---|---|---|---|---|---|---|
| Cg | 1.00 | 0.80 | 0.70 | 0.65 | 0.60 | 0.57 | 0.50 | 0.45 | 0.38 |
Count circuits, not cables and not conductors. A single-phase circuit made of a line conductor, a neutral and a cpc is one circuit. A three-phase circuit is one circuit. A protective conductor never counts, and neither does a cable carrying no load. Counting conductors is the most common error here, and it produces a factor far harsher than the regulation intends — which then gets blamed on the regulation.
The grouping table has other columns for cables laid in a single layer — spaced on a wall or floor, fixed under a ceiling, on perforated tray, on ladder. Those arrangements are kinder, because each cable has air on more sides of it, and a single layer on tray at three circuits is around 0.82 rather than 0.70.
Those columns are not built into this calculator. BS 7671 is not openly published, and the secondary sources available for those columns disagree with one another. This site does not ship a number it cannot verify — the same rule that keeps the reactive component off the voltage drop calculator above 16 mm². Select “other arrangement” and enter Cg from Table 4C1 for your reference method, and the rest of the arithmetic is done for you.
Thermal insulation is extremely good at its job, which is precisely the problem: a cable inside it has no route to shed heat at all. Where a cable is totally surrounded by thermal insulation over a length greater than about half a metre, the capacity is halved — Ci = 0.5, from Regulation 523.9. Shorter surrounded lengths take graded factors from Table 52.2.
This is the factor that most often gets missed, and it is usually nobody’s mistake at the time. A cable clipped along a joist in 1998 was entirely compliant. The loft was topped up with 300 mm of mineral wool in 2015 by someone who had no reason to think about it, and the circuit has been derated by half ever since without anything visibly changing.
A semi-enclosed rewireable fuse is an indecisive device. Its fusing factor is around 2, meaning it can carry roughly twice its marked rating for a long time before the element gives up. The general overload rule assumes a device that operates at 1.45 times its rating, so a cable protected by a rewireable fuse has to be able to survive a much longer overload.
BS 7671 handles it with a single number: It ≥ In ÷ 0.725. And 0.725 is not arbitrary — it is 1 ÷ 1.379, the ratio between the two fusing behaviours. If you are working on a board that still has rewireable fuses, this factor applies to every circuit on it.
Cable selection is two independent tests, and you take whichever demands the larger conductor:
On short runs the current rating usually decides. Past roughly 20–30 m, voltage drop takes over and often demands a cable one or two sizes larger than the current rating alone would. A cable that passes here has not been proved adequate until it has passed there too.
From BS 7671 Appendix 4 or the manufacturer’s data, for your cable type AND your reference method. The installation method matters as much as the conductor size — the same 2.5 mm² cable has a different It clipped direct than it does in an insulated wall.
Ib is what the circuit will carry; In is the nominal rating of the protective device. In must be at least Ib, and the calculator checks the rest.
The air temperature the cable will see in service — a loft or plant room is not 30 °C in August. The insulation type sets the conductor operating temperature the correction is measured against.
Count circuits, not cables and not conductors. If the cables are bunched in conduit or trunking or clipped together, the calculator supplies Cg. For single-layer arrangements on a wall, tray or ladder, read Cg off Table 4C1 and enter it.
A cable totally surrounded by thermal insulation over more than about half a metre takes Ci = 0.5. A BS 3036 rewireable fuse takes Cf = 0.725.
Iz is the capacity you may use, and the calculator confirms whether Ib ≤ In ≤ Iz holds. Current capacity is only one of the two tests — run the same circuit through the voltage drop calculator and take whichever gives the larger conductor.
Iz = It × Ca × Cg × Ci × Cf, where It is the tabulated free-air capacity at 30 °C, Ca is the ambient temperature factor, Cg the grouping factor, Ci the thermal insulation factor and Cf the factor for a BS 3036 semi-enclosed fuse. For sizing you invert it: the tabulated capacity you need is It ≥ In ÷ (Ca × Cg × Ci × Cf), where In is the rating of the protective device. The complete design condition BS 7671 sets is Ib ≤ In ≤ Iz.
Three quantities in order. Ib is the design current — what the circuit will really draw. In is the nominal rating of the protective device. Iz is the derated capacity of the cable where it is installed. The device must be big enough not to trip in normal service (In ≥ Ib) and small enough to protect the cable (In ≤ Iz). Both halves fail in practice: a device below Ib nuisance-trips, and a device above Iz means a sustained overload cooks the cable before anything operates.
Ca = √[(θc − θa) ÷ (θc − 30)], where θc is the conductor’s permitted operating temperature — 70 °C for thermoplastic, 90 °C for thermosetting — and θa is the ambient. The square root comes from the physics: heating goes with I²R, so the current producing a given temperature rise goes with the square root of the headroom available. This one expression reproduces every entry in the published ambient tables to two decimal places, which is why this calculator derives Ca rather than looking it up.
Circuits. A single-phase circuit consisting of a line conductor, a neutral and a cpc is one circuit, not two or three. A three-phase circuit is one circuit. Cables that carry no load do not count, and a protective conductor never counts. Counting conductors instead of circuits is the most common error here and it produces a factor far harsher than the regulation intends.
0.70. The bunched-and-enclosed column runs 1.00 for one circuit, 0.80 for two, 0.70 for three, 0.65 for four, 0.60 for five, 0.57 for six, and on down to 0.38 at twenty. It applies to cables in conduit, trunking or ducting, and to cables bunched and clipped direct to a surface. Single-layer arrangements — spaced on a wall, under a ceiling, on perforated tray or on ladder — use different and generally kinder columns of the same table.
Because we could not verify them to the standard the rest of the site is held to. The bunched column is corroborated by three independent sources and is shipped. For the single-layer columns the secondary sources available disagree with one another, and BS 7671 is not published openly, so there is no way to settle it without the book in hand. Putting a plausible-looking wrong number on a page about conductor sizing would be worse than leaving a gap, so the calculator asks you to enter Cg from Table 4C1 instead and says so on screen.
Where a cable is totally surrounded by thermal insulation over a length of more than about 0.5 m, Regulation 523.9 gives Ci = 0.5 — half the free-air capacity. Insulation is very good at what it does, and a cable inside it has no path to shed heat. This is the factor most often missed, because it usually arrives after the installation: the cable was clipped to a joist and perfectly compliant, and then the loft was topped up. Shorter lengths take graded factors from Table 52.2.
A semi-enclosed rewireable fuse has a fusing factor of about 2, meaning it can carry roughly twice its nominal rating for a long time before clearing. The general overload rule assumes a device that operates at 1.45 times its rating. To keep the cable safe under a fuse that is far less decisive, BS 7671 requires It ≥ In ÷ 0.725 — and 0.725 is simply 1 ÷ 1.379, the ratio between the two fusing behaviours.
No, and conflating the two is a common mistake. Correction factors are about heat: they reduce the current the cable may carry. Voltage drop is about resistance over distance and is unaffected by grouping or ambient in the same way — although the conductor temperature does change its resistance, which is why the voltage drop tables are computed at the insulation’s rated temperature. Size the cable for both tests independently and take whichever demands the larger conductor.
Because It is not a property of the conductor alone — it is a property of the conductor and the installation method together. The same 2.5 mm² twin and earth has one figure clipped direct, a lower one in a wall with insulation on one side, and a lower one again in conduit in an insulated wall. A dropdown offering "2.5 mm² → 27 A" would be confidently wrong most of the time. The tables also sit inside BS 7671, which is not openly published, so reproducing them here is not an option this site will take.
Below about 0.5, the installation conditions rather than the cable are the problem, and upsizing the conductor to compensate gets expensive fast — and often still fails, because a bigger cable in the same hot bunch is still in the same hot bunch. Splitting the group across two routes, moving the run clear of thermal insulation, or improving ventilation will usually cost less than the copper and will fix the cause rather than compensating for it.
Not directly. The NEC arrives at the same place by a different route: Table 310.15(B)(1) for ambient correction and 310.15(C)(1) for more than three current-carrying conductors, then the 110.14(C) terminal temperature limitation on top, which has no BS 7671 equivalent. That whole chain is implemented on the wire size calculator instead of being bolted on here.
Technical reviewer
Electrician · 10+ years of installation work in Bangladesh and the wider South Asian region
He reads the result the way an installer would: are the defaults values people actually meet, does the warning fire where you would stop and think, and is the answer something you could buy and fit? The code figures themselves come from the published standards cited below, not from him — that boundary is set out on his profile.
The full process is written up in the methodology and editorial policy. Results are engineering guidance, not a code sign-off — see the disclaimer. If a result looks wrong, tell us; corrections are answered before anything else.
Voltage drop for DC, single-phase and three-phase runs in copper or aluminium, sized in mm² or AWG, to BS 7671, IEC 60364 or NEC limits — with the BS 7671 Appendix 4 mV/A/m figure shown alongside.
S=√(I²t)/kMinimum conductor size to survive a fault, S = √(I²t) ÷ k, to BS 7671 Regulation 543.1.3 — with k derived from the conductor material and its permitted temperature rise rather than looked up.
AWGWhat size wire you need, from the load, the run length and the conditions — sized against both NEC ampacity and voltage drop, with the terminal temperature rule that stops 90 °C wire giving 90 °C ampacity.
kVASize a house service the way an electrician does: NEC Article 220 standard calculation, with the demand factors that stop a 40 kW house needing a 400 A service.