Containment Fill & the 45% Space Factor Explained

Containment fill is one of the easiest calculations on a project to do and one of the easiest to misread. The arithmetic is a sum of cable areas against a percentage of a box. The trap is what that percentage means: a trunking sized correctly to the 45 % rule looks, on the drawing and on site, as though it is only half used — and the temptation to "use the space that is obviously there" is exactly what the rule exists to stop.

The short version: enclosed containment is limited to 45 % fill for trunking and 35 % for duct (IET Guidance Note 1, Appendix A). That limit is not spare room — round cables cannot fill a section solidly, they have to be installable, and the air is part of the cooling. Conduit is not sized on a percentage at all: GN1 gives every cable a factor and every conduit size a factor, and the conduit's factor falls as the run gets longer and gains bends. Open containment (tray, ladder, basket) is sized on width in a single layer instead, with no space factor. And whichever you use, the number of circuits you have just bunched together sets the grouping factor C_g that shrinks every cable's rating.

Try it: the fill explorer

Add cables until the trunking is at its limit, then watch the two percentages diverge. The "of the whole section" figure is what your eye sees; the "of the permitted fill" figure is what the standard is asking about. Everything here runs on the same engine as the containment fill calculator.

Why 45 %, and where the other 55 % goes

The space factor answers a question that has nothing to do with electricity: how much of a box can cables actually occupy? Three things eat the difference, and none of them is conservatism.

Round cables cannot tile a rectangle. The densest possible packing of identical circles — a perfect hexagonal stack — still leaves about 9 % air, and that is a theoretical best with cables that are all the same size and laid deliberately. A mixed bundle pulled into a trunking on site settles nearer 60 to 70 % of the section at best, and nobody is measuring it as they go.

Cables have to be installable. A trunking filled to its geometric maximum cannot be wired. The allowance is what lets cables be laid in and dressed without being forced past each other, and what lets the next person add a circuit in two years without damaging the insulation of everything already in there.

The air carries the heat away. Cables in an enclosure warm each other, and the free space is part of how that heat gets out. This is the same physics as the grouping factor, which is why filling the containment and derating the cables are two halves of one decision rather than two separate jobs.

A 50 by 50 millimetre trunking holding twelve 10 millimetre cables, next to two bars: the cables occupy 41 per cent of the whole internal section but 91 per cent of the fill the 45 per cent space factor permits. — The same trunking read two ways. Twelve 10 mm cables fill 41 % of the section — and 91 % of what the 45 % space factor allows. A trunking that looks half empty is nearly full.

41 % of the section is 91 % full

Take twelve cables of 10 mm overall diameter. Each has a cross-section of πd²/4 = 78.5 mm², so the twelve need 942 mm². A 50 × 50 mm trunking has roughly 2 300 mm² of usable internal section, of which 45 % — about 1 037 mm² — may be occupied.

So the cables take 942 of the 1 037 mm² they are allowed: 91 % of the permitted fill. They also take 942 of the 2 300 mm² that physically exists: 41 % of the section. Both numbers are correct. Only the first one is the rule, and a calculator that shows you the second is not telling you that you have room.

Four columns comparing fill limits of 100, 60, 45 and 35 per cent, showing the trunking size each would require for the same twelve cables. — Tighten the fill limit and the same cables need a bigger containment. The 45 % is the design rule, not slack left over in the calculation.

Run the same twelve cables against different limits and the point lands: at a notional 100 % they fit a 50 × 50 easily, at 45 % they only just fit it, and at the 35 % that applies to duct they need the next size up. The limit is not describing how full the box looks. It is choosing the box.

Three methods, not one

A space factor is a property of an enclosure, so it does not transfer to tray, ladder or basket. Open containment is normally sized on the width the cables occupy laid out in a single layer — the sum of the outer diameters if they are touching, or the diameters plus the clear gaps if they are spaced at half, one or two diameters.

A trunking cross-section with cables stacked, sized on total cross-sectional area against a 45 per cent space factor, beside a cable tray with cables in a single row, sized on total cable width. — Trunking and duct are sized on area against a space factor. Tray, ladder and basket are normally sized on the width a single layer occupies — a different calculation, not a different number.

Conduit: the one that is not a percentage

Conduit is the type most people size by eye, and it is the one where eye and rule diverge most. GN1 Appendix A gives each cable a dimensionless cable factor and each conduit size a conduit factor; the cables fit if the factors sum to no more than the conduit's. What makes it different from every other method is that the conduit factor is not a property of the conduit alone — it drops as the run gets longer and as bends are added, because the criterion is not "do they fit in the bore" but "can you pull them in without damaging them".

There are two regimes and they use different numbers for the same cable. A run up to 3 m with no bends reads Tables A1 and A2; anything longer, or with any bend at all, reads Tables A3 and A4 — where the same 2.5 mm² conductor is worth a smaller factor and the conduit is worth much less. That is why twelve cables that pass in a 25 mm conduit over a short straight drop need 32 mm once the run is 6 m with two bends. Nothing about the cables changed. If a run pushes you past the tables, the answer is a draw-in box, not a bigger number: splitting the run restarts the length and bend count.

Two details are worth getting right. First, the spacing allowance belongs between cables, not after the last one: five 10 mm cables at 2D spacing need 5 × 10 + 4 × 20 = 130 mm of tray, not 150. Second, if you bunch the cables on the tray rather than laying them out, you are back to an area calculation — and you have also given up the single-layer grouping factors, which is usually the more expensive half of that decision.

The fill is an input to the cable size

This is the step that gets missed. Choosing the containment decides how many circuits are bunched together, and that sets the grouping factor C_g from BS 7671 Table 4C1. C_g multiplies down the current-carrying capacity of every cable in the group, so the containment decision reaches straight back into the cable schedule.

Grouping factor C_g — BS 7671 Table 4C1, items 1–4 (indicative subset)
Arrangement	1	2	3	4	6	9	12	20
1 · Bunched, enclosed (conduit & trunking)	1.00	0.80	0.70	0.65	0.57	0.50	0.45	0.38
2 · Single layer on a wall or floor	1.00	0.85	0.79	0.75	0.72	0.70	0.70	0.70
3 · Single layer on a perforated tray	1.00	0.88	0.82	0.77	0.73	0.72	0.72	0.72
4 · Single layer on a ladder or cleats	1.00	0.87	0.82	0.80	0.79	0.78	0.78	0.78

Circuits per group across the top. Conduit and trunking are not a separate table — they are item 1, "enclosed". Note 2 to the table: where the horizontal clearance between adjacent cables exceeds twice their overall diameter, no factor need be applied at all.

The numbers are blunt. Nine circuits bunched in trunking keeps half of each cable's tabulated rating; the same nine in a single layer on a ladder keep 0.78. And Note 2 is the get-out worth knowing: space the cables by more than twice their diameter and the factor disappears entirely. A wider tray with the cables spaced is frequently cheaper than the larger conductors that bunching would force you into — which is a containment decision made for a cable reason, and the reason the two calculations belong together.

What else is in the trunking

Fill is not the only thing that decides whether cables may share a containment. Regulation 528.1 keeps Band I circuits (ELV, data, signalling) out of the same wiring system as Band II circuits (LV power) unless every Band I cable is insulated for the highest voltage present, or the two are separated by a partition or run in separate compartments. Cables of a fire safety service are dealt with separately again by Section 560, which requires them to be kept apart from other circuits — and their supports have to keep holding them up during a fire, which is BS 8519's territory.

A multi-compartment trunking is the usual answer, and it changes the sum: each compartment is its own fill calculation, not a share of one big section.

Doing it on a real job

Reserve the spare before you choose the size, not after. BS 7671 names no figure; 20–30 % is common on risers and main routes. Setting it aside first means the size you specify has the spare in it rather than being at its limit on day one.
Use the real cable diameters. Overall diameter varies between manufacturers and between an LSOH and a SWA version of the same conductor size — sometimes by enough to change the containment size. Take them from the datasheets for the cables actually specified.
Count the layers. The width method assumes one. If the cables are two deep the single-layer grouping factors no longer apply, and you should be using the bunched figure or a wider containment.
Carry the weight forward. The cables in the containment are a load on the supports, and the support spacing follows from it — see the MEP bracket calculator for the BS 8519 side of that.
Record the assumptions. The space factor used, the spare allowed and the cable data source belong on the calculation, because they are the things a checker will want to see and the things that change between revisions.

Try the numbers on your own route. The containment fill calculator sizes trunking, duct, tray, ladder and basket from a cable schedule, applies the space factor and your spare allowance, reports the C_g your fill implies, and prints a calculation record. Every value in it is indicative and carries a VERIFY note — confirm them against the products actually specified before the calculation is issued.

Frequently Asked Questions

Why is trunking limited to 45% fill?

Because cables cannot use the whole section. Round cables cannot tile a rectangle — even a perfect hexagonal stack of identical circles leaves about 9 per cent air, and a real mixed bundle settles well short of that. The cables also have to be drawn in without forcing them past each other, and the remaining air is part of how the heat escapes. The 45 per cent for trunking and 35 per cent for duct come from IET Guidance Note 1 Appendix A, and they are the design rule, not slack in the calculation. Conduit is the exception: GN1 sizes it by summing a cable factor per cable against a conduit factor for the size, run length and number of bends, which is a drawing-in test rather than a fill percentage.

My trunking passes but looks half empty — is that right?

Yes, and it is the most common misreading of a fill calculation. Twelve 10 mm cables in a 50 by 50 mm trunking occupy about 41 per cent of the internal section, which sounds like plenty of room. But 41 per cent of the section is 91 per cent of what the 45 per cent space factor permits. The picture and the percentage are answering two different questions, and only one of them is the rule.

Does the space factor apply to cable tray and basket?

Not normally. A space factor is a fill limit for an enclosed system. Open containment is sized on the width the cables occupy in a single layer: the sum of the outer diameters for cables laid touching, or the diameters plus the clear spacing for cables laid half, one or two diameters apart. The area method returns only if you bunch the cables rather than laying them out — at which point the single-layer grouping factors stop applying too.

How much spare capacity should I leave?

BS 7671 gives no figure — it is a project decision, and 20 to 30 per cent is common on risers and main routes where cables get added later. Reserve it before you choose the size rather than after, so the containment you specify has the spare set aside instead of being at its limit on the day it is installed. Many employer's requirements name a percentage, so check the specification first.

Does filling the containment change the cable size?

Yes, and this is the step that gets skipped. The number of circuits sharing the containment sets the grouping factor Cg from BS 7671 Table 4C1, and Cg multiplies down every cable's current-carrying capacity. Nine circuits bunched in trunking keeps only half. Cable size and containment size are not two separate calculations — the fill you choose is an input to the cable sizing, and a bigger containment with the cables spaced can be cheaper than the larger cables that bunching would demand.

Can power and data cables share a trunking?

Only under conditions. Regulation 528.1 requires that Band I circuits (ELV, data, signalling) and Band II circuits (LV power) are not placed in the same wiring system unless every Band I cable is insulated for the highest voltage present, or the two are separated by a partition or run in separate compartments. Cables of a fire safety service are governed separately by Section 560, which requires them to be kept apart from other circuits — and their supports have to survive a fire, which is where BS 8519 comes in.

Sources and verification. IET Guidance Note 1 Appendix A (space factors and the trunking capacity tables); BS 7671:2018+A4:2026 Regulation 528.1 (Band I and Band II segregation), Section 560 and Regulation 560.7.1 (fire safety services), Appendix 4 Table 4C1 items 1–4 and its Note 2, and Table 4C4 (spaced tray and ladder). Cable dimensions are indicative catalogue values. Confirm every figure against the products actually specified before a calculation is issued.