Cable Derating Factors Explained: Grouping, Ambient & Installation

The current rating you read from a cable table is a best case — one circuit, in free air, at a comfortable temperature, protected by a device that trips near its rating. Real routes are hotter and more crowded than that, so the rating comes down. What catches people out is not any single factor. It is that they multiply: three conditions that each look survivable can leave you with less than half the cable you thought you had.

The short version: the tabulated capacity I_t assumes reference conditions, and every way your job differs costs a factor. They multiply, so 0.94 × 0.65 × 0.725 = 0.443 — 44 % of the cable. Divide the device rating I_n by the product, not the load, because Regulation 433.1.1 requires I_b ≤ I_n ≤ I_z. And know the get-outs: spacing cables beyond twice their diameter removes the grouping factor entirely.

Try it: the derating explorer

Start with a cable that comfortably passes, then add one condition at a time. Watch the capacity fall and — if you pick a rewireable fuse or bury the cable — watch the requirement rise to meet it. The factors, the tables they come from and the smallest size that would still pass are all live.

Why derating exists: it is all about heat

A cable's rating is simply the current at which its conductor reaches its maximum safe temperature — 70 °C for thermoplastic (PVC), 90 °C for thermosetting (XLPE). Anything that makes it harder for the cable to shed that heat lowers the current it can safely carry. Every rating factor is a different way of saying "this cable cannot lose heat as easily as the table assumed", so they are not arbitrary penalties. They are the heat-transfer physics of the real installation.

That is also why a factor can be greater than 1. A cable in a genuinely cold environment sheds heat more readily than the table assumed, and Appendix 4 lets you have the difference.

What the tabulated rating quietly assumes

A rating factor is not a penalty. It is the difference between the installation the table was calculated for and the one you are actually building.

Appendix 4 §2.1 sets the reference ambient temperatures: 30 °C for cables in air, whatever the installation method, and 20 °C for cables buried directly in soil or in a duct in the ground. The tables also assume one circuit on its own, a cable free to lose heat to its surroundings, and a protective device that operates near its rating. Break none of those and every factor is 1.00.

Appendix 4 §3 names the whole family, and it is worth knowing all seven letters rather than the usual three:

C_a — ambient temperature (Table 4B1 air, Table 4B2 ground)
C_g — grouping (Tables 4C1 to 4C6)
C_i — thermal insulation (Reg 523.9, Appendix 4 §2.6)
C_f — protection by a semi-enclosed fuse to BS 3036 (Reg 433.1.202)
C_s — soil thermal resistivity (Table 4B3)
C_d — depth of burial (Table 4B4)
C_c — buried circuits protected against overload (Reg 433.1.203)

Divide the device rating, not the load

This is the most common error on the page, and the standard is unusually direct about it. Regulation 433.1.1 requires I_b ≤ I_n ≤ I_z — the cable has to clear the protective device, not merely the load. So the sizing runs backwards from I_n:

I_t ≥ I_n / (C_g · C_a · C_s · C_d · C_i · C_f · C_c)

That is Appendix 4 Equation 2. Section 5 explains why it is written as a division rather than a multiplication: the ambient, grouping and insulation factors could be applied to I_t as multipliers if you preferred, but the BS 3036 factor "has to be applied to I_n as a divisor", so "it is therefore more convenient to apply all the rating factors to I_n as divisors".

The practical consequence: a 25 A load on a 32 A breaker is sized on 32 A. Dividing the design current instead of the device rating quietly undersizes the cable by whatever headroom you left when you picked the device — and headroom is exactly what most people leave.

They multiply, and that is brutal

A bar chart of a 6 square millimetre cable rated 46 amps losing capacity step by step: an ambient factor of 0.94, then a grouping factor of 0.65, leaving 28.1 amps — while a BS 3036 fuse raises what the cable must clear from 32 amps to 44.1 amps, so the cable fails. — Three individually unalarming conditions leave 44 per cent of the cable. Note that the capacity falls while the requirement rises: Appendix 4 applies the fuse factor to the device rating as a divisor, not to the cable as a multiplier.

Take a 32 A circuit on 6 mm² PVC multicore, clipped direct — Reference Method C, tabulated at 46 A. Now give it three ordinary conditions:

Worked example. A 35 °C plant room gives C_a = 0.94. Four circuits bunched together give C_g = 0.65. The board is old and the circuit is on a BS 3036 rewireable fuse, so C_f = 0.725.
Capacity: 46 × 0.94 × 0.65 = 28.1 A
Requirement: 32 ÷ 0.725 = 44.1 A
The cable is short by 16 A. Overall the factors multiply to 0.443, so 44 % of the tabulated capacity survives — and a circuit you would have run in 6 mm² needs 16 mm².

None of those three conditions is unusual, and none looks alarming written down. 0.94 is a 6 % reduction. 0.65 sounds like a third. 0.725 sounds like a quarter. The arithmetic itself is not really the surprise — the surprise is how ordinary the conditions are, and how rarely all three get written down on the same drawing.

C_a — ambient temperature

Two tables, and using the wrong one is a real error rather than a rounding difference. Table 4B1 is ambient air on a 30 °C base; Table 4B2 is ambient ground on a 20 °C base. A buried cable read off the air table is handed a rating it has not earned, because 4B1 has no row below 25 °C to give it.

Table 4B1 — rating factors for ambient air temperatures other than 30 °C (indicative subset)
Cable	25 °C	30 °C	35 °C	40 °C	45 °C	50 °C	55 °C	60 °C
70 °C thermoplastic (PVC)	1.03	1.00	0.94	0.87	0.79	0.71	0.61	0.50
90 °C thermosetting (XLPE)	1.02	1.00	0.96	0.91	0.87	0.82	0.76	0.71

In the ground the equivalent 4B2 values for a 70 °C thermoplastic cable run 1.10 at 10 °C, 1.00 at the 20 °C base, 0.89 at 30 °C and 0.77 at 40 °C. Note that Appendix 4 §2.1 lets you ignore a soil temperature that exceeds the selected ambient by up to 5 °C for only a few weeks a year, and that neither table accounts for solar gain.

C_g — grouping, and the one most tools get wrong

Line chart of the grouping factor against the number of circuits for cables in air. Bunched or in trunking falls to 0.38 at twenty circuits; a single layer on a wall reaches 0.70, on a perforated tray 0.72 and on a ladder 0.78. The Table 4C4 spaced rows stop at six cables: spaced on a tray holds 0.91 and spaced on a ladder stays at 1.00. — The same cables, differently arranged. Bunching — which is what conduit and trunking are — is the worst case and the one most tools assume. Spacing costs almost nothing: 20 mm between cables on a ladder leaves the factor at 1.00, where bunching six of them costs 43 per cent.

Table 4C1 has four rows, not one. Almost every calculator — including, until recently, ours — applies item 1, "bunched in air, on a surface, embedded or enclosed", because it is the worst case and therefore the safe assumption. But if your cables are in a single layer on a perforated tray, the standard gives you a materially better factor, and taking the bunched value means specifying copper you do not need.

BS 7671 Table 4C1 — rating factors for groups of more than one circuit or multicore cable
Arrangement (cables touching)	1	2	3	4	5	6	7	8	9	12	16	20
1 · Bunched, enclosed or on a surface	1.00	0.80	0.70	0.65	0.60	0.57	0.54	0.52	0.50	0.45	0.41	0.38
2 · Single layer on a wall or floor	1.00	0.85	0.79	0.75	0.73	0.72	0.72	0.71	0.70	0.70	0.70	0.70
3 · Single layer on a perforated tray	1.00	0.88	0.82	0.77	0.75	0.73	0.73	0.72	0.72	0.72	0.72	0.72
4 · Single layer on a ladder or cleats	1.00	0.87	0.82	0.80	0.80	0.79	0.79	0.78	0.78	0.78	0.78	0.78

Those are the columns the table actually prints. Between them, round the circuit count up: ten circuits takes the twelve-circuit column. Conduit and trunking are not a separate table — they are item 1, "enclosed", which is why you will never find a trunking row however hard you look. Groups of single-core cables have their own table (4C5) with different values again.

Spacing is the cheapest fix on this page

Table 4C1 assumes the cables are touching. Table 4C4 covers the same trays and ladders with the cables spaced, and the difference is remarkable: 20 mm between cables on a ladder leaves the factor at 1.00 however many you put on it, where bunching six of them costs 43 %. Twenty millimetres of air is almost always cheaper than a cable size.

BS 7671 Table 4C4 — a single tray or ladder with the cables spaced (cables per tray)
Arrangement	1	2	3	4	6
Perforated tray, spaced ≥ 20 mm	1.00	1.00	0.98	0.95	0.91
Vertical tray, spaced 225 mm	1.00	0.91	0.89	0.88	0.87
Ladder or cleats, spaced ≥ 20 mm	1.00	1.00	1.00	1.00	1.00

Tabulated to six cables per tray only. Table 4C4 also covers multiple trays or ladders stacked above one another, which reduces the factor again — 300 mm vertical spacing between trays is assumed, and closer spacing than that needs a lower value still.

When the grouping factor need not apply at all

These are worth as much as the factors themselves, and are almost never explained:

Clearance — Table 4C1 NOTE 2. "Where horizontal clearances between adjacent cables exceed twice their overall diameter, no rating factor need be applied." Spacing cables out is frequently cheaper than the size up that bunching them would cost.
The 30 % rule — Reg 523.5 and NOTE 9. A cable expected to carry no more than 30 % of its grouped rating "may be ignored for the purpose of obtaining the rating factor for the rest of the group". A group of six that includes two lightly loaded circuits may only be a group of four.
Only over the grouped length. The factor applies where the cables are actually bunched. A run that is grouped for a short section and open for the rest may not be governed by that section — though you have to be able to justify it.
Mixed operating temperatures. Where cables of different maximum operating temperature are grouped, the rating for all of them is based on the lowest. One PVC cable in a group of XLPE drags the whole group down.

C_i — the thermal-insulation cliff

Regulation 523.9 is blunt: a cable "should preferably not be installed in a location where it is liable to be covered by thermal insulation". Where it must be, and is totally surrounded for 0.5 m or more, the capacity is taken as 0.5 times the Reference Method C value, in the absence of more precise information. For shorter lengths Appendix 4 §2.6 gives:

Appendix 4 §2.6 — cable totally surrounded by thermal insulation for less than 0.5 m
Length in insulation	50 mm	100 mm	200 mm	400 mm	Over 500 mm
Factor C_i	0.88	0.78	0.63	0.51	0.50

These apply to conductors up to 10 mm² in insulation with a thermal conductivity above 0.04 W m⁻¹K⁻¹. A cable in an insulated wall touching a thermally conductive surface on one side is handled differently — through Reference Method A rather than through a factor.

C_f — the BS 3036 fuse

A semi-enclosed rewireable fuse carries a great deal more than its nominal rating before it goes. Appendix 4 §4 derives the correction as 1.45 / 2 = 0.725, a value which "results in the same degree of protection as that afforded by other overload protective devices", and says it is applied "to the nominal rating of the fuse as a divisor". Regulation 433.1.202 states the same requirement from the other end: I_n must not exceed 0.725 × I_z.

Rewireable fuses are rare in new work but common in the boards you find on site, and the factor is severe enough to change the answer on its own — on a plain 32 A circuit it is the difference between 6 mm² and 10 mm².

Buried cables: C_s, C_d and C_c

A buried cable is not the same calculation with two extra factors bolted on — four things change at once, and the last of them is the one that gets forgotten.

The ambient table changes from 4B1 (air, 30 °C base) to 4B2 (ground, 20 °C base).
C_s — soil thermal resistivity, Table 4B3. The reference is 2.5 K·m/W. Direct-buried cables are far more sensitive than cables in ducts: at 1.0 K·m/W direct burial gives 1.50 against a duct's 1.18, and at 3.0 it gives 0.90 against 0.96.
C_d — depth of laying, Table 4B4. The reference is 0.7 m. Depth matters less than people expect: 2 m only costs about 0.92.
C_c = 0.9 — Reg 433.1.203. Where a buried circuit is protected against overload, I_n must not exceed 0.9 × I_z. It does not look like a rating factor and it is not in the 4B or 4C tables, which is exactly why it gets missed.

Buried grouping is a different table entirely

This is the part that surprises people. Tables 4C2 and 4C3 do not ask how the cables are arranged — they ask how far apart they are. A buried group read off the in-air bunched row is being answered with the wrong question.

Line chart of the buried grouping factor against the number of circuits or ducts. Direct in the ground and touching falls to 0.50 at six circuits; at 250 mm clearance it holds 0.70 and at 500 mm 0.80. Cables in ducts start higher, and ducts a metre apart hold 0.90. — Buried cables are grouped by cable-to-cable clearance, not by arrangement — a different table and a different idea. Two circuits touching in the ground cost 0.75; the same two half a metre apart cost only 0.90, which is usually a trench width rather than a cable size.

BS 7671 Tables 4C2 (buried direct) and 4C3 (in buried ducts) — by number of circuits or ducts
Clearance	2	3	4	5	6
Direct in ground, touching	0.75	0.65	0.60	0.55	0.50
Direct, one cable diameter	0.80	0.70	0.60	0.55	0.55
Direct, 125 mm apart	0.85	0.75	0.70	0.65	0.60
Direct, 250 mm apart	0.90	0.80	0.75	0.70	0.70
Direct, 500 mm apart	0.90	0.85	0.80	0.80	0.80
In ducts, touching	0.85	0.75	0.70	0.65	0.60
In ducts, 250 mm apart	0.90	0.85	0.80	0.80	0.80
In ducts, 500 mm apart	0.95	0.90	0.85	0.85	0.80
In ducts, 1 m apart	0.95	0.95	0.90	0.90	0.90

Both tables assume 0.7 m depth and 2.5 K·m/W soil. Ducts start higher than direct burial because the air gap around the cable already helps. Half a metre between trench runs takes six touching circuits from 0.50 to 0.80 — usually a trench-width decision rather than a cable-size one.

C_n — more than four loaded cores

Table 4B5 handles a multicore carrying more than four loaded cores — and the word loaded matters, because a protective conductor does not count. Control and instrument multicores hit this constantly and it is rarely applied:

BS 7671 Table 4B5 — rating factors for cables having more than 4 loaded cores
Loaded cores	5	6	7	10	12	14	19	24	30	37	48
Factor	0.72	0.67	0.63	0.56	0.53	0.51	0.45	0.42	0.39	0.36	0.33

Table 4B5 NOTE 1 scopes this: for cables of 1.5 to 4 mm² the capacity is the two-core value for the same insulation type multiplied by the factor above.

C_h — triplen harmonics

Third-harmonic currents from electronic loads do not cancel in the neutral of a three-phase circuit the way the fundamental does — they add. Appendix 4 §5.5 and Table 4Aa handle it in two stages, and the second stage is the one that catches people:

Below 15 % third harmonic — no factor. Size on the line current.
15 % to 33 % — factor 0.86, still on the line current.
33 % to 45 % — factor 0.86, but now size on the neutral current.
Above 45 % — no factor, but size on the neutral current, which by this point exceeds the line current.

Above 33 % the answer is no longer a factor at all — it is a change of which current you size on, and the neutral can end up carrying more than any line does. On a floor of switched-mode supplies that is a real design condition, not a curiosity.

If a cable fails the check

Space the cables out — beyond twice their overall diameter the grouping factor disappears entirely. Usually the cheapest fix available.
Change the arrangement — a single layer on a tray or ladder instead of bunched is worth a great deal: at six circuits, 0.79 on a ladder against 0.57 bunched.
Recount the group — apply the 30 % rule and see whether the lightly loaded circuits belong in the count at all.
Move the route out of the hot part, or out of the insulation. C_i is a cliff rather than a slope.
Check the reference method — clipped direct rather than in conduit in an insulated wall starts you on a much higher column before any factor applies.
Replace a BS 3036 fuse with a modern device and recover 27.5 % at a stroke.
Then increase the conductor size — the answer, but the last one to reach for, because every other option is cheaper.

Where this fits in a cable calculation

Derating is the current-carrying-capacity check — the first of the five checks in a compliant cable calculation, alongside voltage drop, earth-fault loop impedance, the adiabatic check on the protective conductor, and breaking capacity against the prospective fault level. It is also the check most likely to be quietly wrong, because nothing on site tells you a cable is running hot until it has been doing it for years.

Two related traps worth reading next: a 90 °C cable terminated into 70 °C equipment loses its headline rating before any factor is applied, and the choice between PVC, XLPE and SWA changes which table you start from. Get the design current itself from maximum demand and diversity, or straight from a load in kilowatts with the kW to amps calculator.

Our free BS 7671 cable sizing calculator applies all of these factors across a whole distribution board and prints a calculation report with the arithmetic shown.

Glossary

I_t — the tabulated current-carrying capacity, read from Appendix 4 before any factor is applied.

I_z — the effective capacity of the cable as installed: I_t after the factors.

I_b — design current: what the circuit is actually intended to carry, after diversity.

I_n — the rated current of the protective device, and what the rating factors divide.

Reference method — the installation arrangement (A, B, C, E, F…) that decides which capacity column you start from.

Reference conditions — 30 °C in air, 20 °C in ground, one circuit, no thermal insulation: what the tables assume.

Bunched — Table 4C1 item 1: cables touching, enclosed or on a surface. The worst case, and the usual default.

K·m/W — thermal resistivity: how strongly soil resists carrying heat away. Higher is worse for the cable.

Semi-enclosed fuse — a BS 3036 rewireable fuse, the one that attracts the 0.725 factor.

Frequently Asked Questions

Do the derating factors add or multiply?

They multiply. The required tabulated rating is the device rating divided by the product of every applicable factor, so several modest factors combine into a large reduction. A 35 °C ambient (0.94), four circuits bunched (0.65) and a BS 3036 fuse (0.725) multiply to 0.443 — the cable keeps only 44 per cent of its tabulated capacity.

Do you derate the design current or the device rating?

The device rating, In. Regulation 433.1.1 requires Ib ≤ In ≤ Iz, so the cable has to clear the protective device, not merely the load. Appendix 4 §5 says it plainly: "it is therefore more convenient to apply all the rating factors to In as divisors". A 25 A load on a 32 A breaker is sized on 32 A, and dividing by the load instead is a common way to undersize a cable.

What is the grouping factor for four cables touching?

It depends entirely on how they are arranged. BS 7671 Table 4C1 gives 0.65 for four circuits bunched, enclosed or on a surface; 0.75 in a single layer on a wall or floor; 0.77 in a single layer on a perforated tray; and 0.80 on a ladder. Most calculators assume the bunched figure because it is the worst case, so a tray or ladder run is often derated far harder than the standard requires.

When does the grouping factor not apply?

Three cases. Table 4C1 NOTE 2: where the horizontal clearance between adjacent cables exceeds twice their overall diameter, no factor need be applied at all. Regulation 523.5 and NOTE 9: a cable expected to carry no more than 30 per cent of its grouped rating may be left out of the count entirely. And the factor only has to be applied over the length where the cables are actually bunched, so a short grouped section may not govern the whole run.

Does a cable in thermal insulation really derate to 0.5?

Yes, where it is totally surrounded for 0.5 m or more. Regulation 523.9 says the current-carrying capacity shall be taken, in the absence of more precise information, as 0.5 times the capacity for that cable clipped direct to a surface and open — Reference Method C. For shorter lengths Appendix 4 §2.6 gives 0.88 at 50 mm, 0.78 at 100 mm, 0.63 at 200 mm and 0.51 at 400 mm, for conductors up to 10 mm².

Can a cold environment increase a cable's rating?

Yes. The ambient factor works both ways: below the reference temperature it is greater than 1. In air the reference is 30 °C, so 25 °C gives 1.03 for a 70 °C thermoplastic cable. In the ground the reference is 20 °C and 10 °C gives 1.10. The low temperature has to be reliable year-round to be worth claiming, and a buried cable must use the ground table, not the air one.

Why is there a 0.725 factor for a BS 3036 fuse?

Because a semi-enclosed rewireable fuse can carry a great deal more than its nominal rating before it operates. Appendix 4 §4 derives the factor as 1.45 / 2 = 0.725, "which results in the same degree of protection as that afforded by other overload protective devices". Regulation 433.1.202 puts the same requirement the other way round: In must not exceed 0.725 times the current-carrying capacity of the lowest rated conductor in the circuit.

What extra factors apply to a buried cable?

Four things change. The ambient comes from Table 4B2 (ground, 20 °C base) rather than Table 4B1 (air, 30 °C). Table 4B3 gives Cs for soil thermal resistivity other than 2.5 K·m/W. Table 4B4 gives Cd for a depth of laying other than 0.7 m. And Regulation 433.1.203 adds Cc = 0.9 wherever the circuit is protected against overload — a factor that is easy to miss because it does not look like a rating factor at all.

Which derating factor is missed most often?

Hidden grouping — a bunched section that is not obvious on the drawing — and the thermal-insulation factor where a cable passes through an insulated wall or a loft. Both reduce the safe current and both are easy to overlook, which is the dangerous direction of error. Running a close second is using the air ambient table for a buried cable, which quietly hands the cable a rating it has not earned.

Sources and verification. BS 7671:2018+A4:2026 Regulations 433.1.1, 433.1.202, 433.1.203, 523.5 and 523.9; Appendix 4 §2.1 (reference conditions), §2.3 (groups), §2.6 (thermal insulation), §3 (the C factors), §4 (overload protection and the 0.725 derivation), §5 (Equations 1 and 2) and §5.5 (harmonics); Tables 4B1, 4B2, 4B3, 4B4, 4B5, 4C1 to 4C6 and 4Aa. The values on this page are an indicative subset for explanation — verify every figure against your own copy of BS 7671 for the actual cable, installation method and reference table.