Cable Tray Fill Calculation: NEC Limits and Sizing

Cable tray fill calculation describes the area occupied by cables. It needs a separate check against the applicable cable and tray rules, and another against the space needed for bends, access and later changes. A low percentage alone does not establish that an installation complies or that an individual cable remains removable.

For your own cable count and pathway dimensions, use the calculator. This article explains the limits and the rack layout decisions behind those numbers.

NEC 392.22: why there is no universal tray fill percentage

Forty percent can refer to different rules. Chapter 9, Table 1 of NFPA 70 uses it for conduit and tubing containing more than two conductors. Cable tray has separate provisions in Article 392; copying the conduit percentage does not select the correct tray rule.

For example, Legrand’s Cablofil fill reference distinguishes the control/signal case in open-bottom tray at 50 percent under 392.22(A)(2) from solid-bottom tray at 40 percent under 392.22(A)(4). Power cable cases use area tables or diameter-based rules instead. Thus 40 percent is not exclusively a conduit number, nor a universal tray limit.

Eaton’s cited manual explains the 2014 NEC cases on pages MAN-23 and MAN-24. The control/signal calculation caps the credited loading depth at 6 inches, even with taller side rails. Treat that edition-specific explanation as a reference and confirm the adopted NEC edition and product instructions for an actual installation.

Identify the cable system before choosing a limit

The category printed on a data cable does not identify every applicable installation rule. Communications, optical fibre, control circuits and cables delivering remote power require their own classification and references. Do not treat the Chapter 8 structure as a blanket exemption for anything plugged into a network rack.

Record these inputs before selecting the allowable fill:

• Cable listing and circuit use: include power, control, communications and fibre separately.
• Tray construction and loading dimensions: distinguish ladder, trough, basket and solid bottom; use the manufacturer’s classification.
• Project requirements: identify the applicable code edition, manufacturer fill guidance, load rating and reserve for later additions.

The 25 and 40 percent columns below are planning scenarios. They do not grant permission to use those percentages for every cable system. The smaller design target simply reserves more area for growth.

How the percentage is actually computed

Fill is the total cable cross-sectional area divided by the usable cross-sectional area of the pathway.

For one round cable of outer diameter d, the area is pi times (d/2) squared. For n identical cables, multiply by n. The tray’s usable area is its inside width multiplied by the loading depth, which is the depth you intend to fill, not the full height of the side rail.

Worked example, 96 runs of Cat6a at 7.4 mm outer diameter in a 300 mm by 100 mm wire basket:

• Area of one cable: pi x 3.7² = 43.0 mm²
• Total cable area: 43.0 x 96 = 4,129 mm²
• Tray area: 300 x 100 = 30,000 mm²
• Fill: 4,129 / 30,000 = 13.8 percent

Two things about that result. Outer diameter is squared, so it dominates everything: moving from Cat6 at 6.2 mm to Cat6a at 7.4 mm raises the area per cable by 42 percent for the same run count, and shielded constructions run larger again. Take the diameter from the datasheet of the cable being bought, never from a category label. And the loading depth is a choice, so quietly assuming the full side height of a 100 mm basket is how a design ends up at 20 percent on paper and unusable in the rack.

Fill chart for common tray sizes

Counts below are pure geometry, rounded down: usable area times the fill fraction, divided by the area of one cable. The example diameters are 6.2 mm and 7.4 mm; they are assumptions, not category-wide specifications.

Tray, width x depth	Cross section	Cat6 at 40%	Cat6a at 40%	Cat6a at 25%
150 x 50 mm	7,500 mm²	99	69	43
300 x 50 mm	15,000 mm²	198	139	87
300 x 100 mm	30,000 mm²	397	279	174
450 x 100 mm	45,000 mm²	596	418	261
600 x 100 mm	60,000 mm²	795	558	348

The 25 percent column illustrates reserving area for later additions. Select a project target after identifying the applicable limit; neither column includes cable weight, bend space or remote-power heating.

The layer test, and why area fill overstates capacity

Now the part the percentage hides. A tray at 40 percent area fill is not 40 percent full in any sense a technician would recognise. Round cables do not settle into a neat rectangle; they heap into layers, and every cable below the top layer is trapped by the ones above it. If a cable has to stay individually removable, capacity is set by how many layers you allow, not by area:

Tray width	Cat6 per layer	Cat6 at 2 layers	Cat6a per layer	Cat6a at 2 layers
150 mm	24	48	20	40
300 mm	48	96	40	80
450 mm	72	144	60	120
600 mm	96	192	81	162

Compare the two tables at 300 by 100 mm. Area fill at 40 percent gives 279 Cat6a runs. Two straight layers give 80. The difference shows how layout assumptions change the answer; it does not establish a universal maintainable capacity.

One or two layers can be a useful access target, but straps, crossovers and drop-outs also affect removal. Area fill is only one input to a capacity decision. Check load ratings separately and plan how a cable would be released without disturbing unrelated connections.

What binds before the percentage does

Bend radius at every drop-out

A tray run is straight and forgiving. The point where cable leaves the tray and turns down into a rack is neither. The TIA-568 series sets minimum bend radius for balanced twisted-pair cabling as a multiple of the cable’s outer diameter, tighter once installed than during a pull under tension, with a larger multiplier for multi-pair backbone cable than for four-pair horizontal cable. Optical fibre carries its own limits, unloaded and under load.

At the drop-out this becomes a volume problem, not a percentage. A dense bundle leaving a tray needs a radiused exit and enough vertical space below the tray to make the turn without the cables on the inside of the curve going tighter than their limit, and that space is frequently what runs out first.

Use each cable manufacturer’s installed and pulling limits. ITU-T G.657 describes bend-insensitive fibre characteristics, but does not remove the finished cable’s handling limits. Use loose, reusable hook-and-loop straps as recommended in NVIDIA’s cabling guidance; tightening a bundle to force it through an undersized exit is not a sizing method.

Weight and span

Cable tray is rated by load and span together, and the NEMA VE 1 designations state both: a span in feet and a working load class in pounds per foot. The rating applies to the support spacing actually installed, not to the one in the catalogue photograph.

Multiply the cable manufacturer’s mass per unit length by the count, then compare with the tray’s load table at the installed support span. Do not assume that a data-only load is light enough. Fittings and support arrangements also matter, as explained in cable basket versus ladder rack versus solid tray.

Heat from remote power

Data cable stopped being a cold pathway when Power over Ethernet grew. IEEE 802.3bt defines Type 3 and Type 4 powering, with Type 4 sourcing up to 90 W at the power sourcing equipment. Current in a bundle heats its centre, and the rise scales with bundle size because interior cables have no path to ambient.

Consult the applicable NEC remote-power provisions and TIA’s TSB-184 guidance alongside the cable manufacturer’s PoE recommendations. A tray of unpowered horizontal cable and a tray of Type 4 PoE runs need different thermal checks even at identical area fill. Geometry alone cannot establish allowable current or temperature rise.

Rack layout, service loops and labeling

Set patch panels and managers before ordering leads

The fundamentals of the rack belong in the tray plan. Use a patch panel for permanent structured runs and short patch leads for equipment connections. Record panel positions and the route through horizontal and vertical managers before selecting lead lengths. This keeps a port change in the patch field instead of requiring the permanent run to be rearranged.

Schneider Electric’s White Paper 203 connects this layout work with cable entry, routing and documentation. Plan the path from tray drop-out to panel, then from panel to equipment. Keep enough room for connectors and manager covers as well as the cable jacket. For a complete installation sequence, see how to cable manage a server rack.

Keep airflow and power separation in the layout

Route bundles along the rack’s management channels and keep equipment airflow paths clear. Follow the equipment’s actual airflow direction and clearance instructions; front-to-back cooling is not a property of every device. NVIDIA’s guide also calls for access to fan units and transceivers, and support that removes strain from connectors.

Schneider’s rack guidance separates power and data routes and uses right-angle crossings where needed. It also gives fibre its own slack management. Those rack practices do not replace the electrical separation rules for a shared power-and-data tray. Avoid placing fragile fibre beneath heavy copper bundles or storing spare leads against exhaust vents.

Store slack and keep the record current

Allow the service movement specified for a sliding device and its cable management arm. Store the required slack in an appropriate manager or spool with the cable’s bend radius preserved. Excess length should not hang across equipment or turn into a tight coil behind the rack.

Label both ends and keep an endpoint record showing the rack, panel, port and destination. Update it when a lead moves. Schneider’s documentation step makes that record part of the installation, and it keeps later tracing from depending on pulling a cable to find its far end. Review unused leads during changes so abandoned cable does not consume the reserve planned into the tray.

Sizing a tray you will still like in five years

Work in this order rather than starting from a percentage.

Count the eventual terminated positions and take outer diameter from the cable datasheet. Set an access target, calculate area fill, and compare it with the applicable limit. Reserve capacity for later additions, including the required space in managers and at drop-outs. This makes the width decision traceable to the planned installation.

Then check the three constraints above against the specific run: the drop-out geometry at each rack, the support spacing against the load class, and whether the cable will be carrying remote power. To work out how thick a given bundle will be before it reaches the tray, see cable bundle diameter.

FAQ

Q: Does the NEC 40 percent fill rule apply to cable tray?

There is no universal 40 percent tray rule. The conduit table and the solid-bottom control/signal tray case both use 40 percent in different contexts. Article 392.22 selects limits by cable and tray type; verify the applicable case before comparing a fill result.

Q: What fill percentage should a data cabling tray be designed to?

Choose a project target below the applicable limit with enough room for growth and access. This guide uses 25 percent as an illustrative initial target. It is not a universal NEC requirement or a guarantee of serviceability.

Q: How many Cat6a cables fit in a 300 mm cable tray?

By cross-sectional area at a 40 percent planning target, a 300 by 100 mm tray gives 279 runs of 7.4 mm cable. Two straight layers give 80. Both are geometric scenarios; cable listing, tray loading, bends and access still need checking.

Q: Does Power over Ethernet change how full a tray can be?

It adds a thermal check that the area percentage does not provide. Review cable construction, bundle count, ambient temperature and remote-power guidance before deciding how to group powered data cables.