Supporting Cable Tray: Load, Spans and the Room It Feeds

Width is a fill dimension. Nobody writes down the other one

The note on a containment layout reads like a purchase order: 600 mm ladder, hot-dip galvanised, supports at 3 m. Every word of that describes the tray. Not one of them describes what will be lying in it in eighteen months. Width buys cross-sectional area, and cross-sectional area answers a fill question. The threaded rod, the channel and the beam clamp above the tray are answering a completely different question, in kilograms per metre, and that number appears on no drawing anybody hands to the crew.

It also arrives in instalments. First fix drops in the feeders that were on the schedule at tender. Controls add a bundle in month four. The fire alarm contractor runs a loop along the same route because the route is already there. Somebody straps two 32 mm conduits to the outside rail because the rail was convenient. Each addition is defensible on its own and none of them triggers a recalculation, because the person adding cable is not the person who chose the hanger spacing and neither of them is holding the tray manufacturer's load/span table. The whole failure mode of a tray run is that its loading is cumulative, distributed across trades, and invisible until a support lets go — usually the one nearest a bend, usually while somebody is on a lift underneath it.

Where the kilograms come from

Start with the cable schedule and nothing else, because it is the only contributor anyone has published numbers for. Every cable type has a mass per unit length in its manufacturer's data sheet, and the honest way to build the total is to list the cables in this stretch of tray and add their figures. Resist the urge to estimate from conductor size: armouring, bedding, sheath material and whether the cable is single-core or multicore move the mass per metre far more than the copper does, and two cables of the same current rating can differ by half again.

Then the tray itself, from the catalogue for the section actually delivered. Aluminium ladder is light enough to carry up a ladder in three-metre lengths; hot-dip galvanised steel solid-bottom of the same nominal width is several times that, and a fibreglass system to NEMA FG 1 is heavier than its stiffness suggests. Covers are the item most often left out of the sum entirely — they are bought as an accessory, invoiced separately, and weigh what they weigh whether or not the load calculation remembers them.

After that come the things nobody calls tray load. Small conduits clipped to the side rail. Single-core cleats. A pneumatic line that shares the route because the route exists. An earth tape. A festoon of draw wires left in from the last pull. Individually each is trivial and that is precisely why they accumulate — no one item is worth a calculation, and the sum of them is a real fraction of the cable weight on a lightly loaded run.

Outdoors, add the weather. Radial ice accretes on the tray and on every cable in it, and the accreted mass is governed by the exposed surface rather than by the cable weight, so a lightly loaded outdoor run can gain proportionally more than a heavily loaded one. ASCE/SEI 7 covers atmospheric ice loading, and the local building code names the edition and the site's design ice thickness; do not carry a figure from the last project across a climate boundary. Cold-store and canopy runs are the two that catch people out, because they read as indoor work right up until the frost forms.

Do it per segment, because the schedule thins out

A route that leaves a switchroom with twelve feeders in it and reaches the far end with three is not one loading case wearing a single label, and sizing the whole thing on the heaviest stretch buys steel and hangers that nobody needs. Break the run at every branch point and treat each stretch between branches as its own problem. Risers are the sharpest version of this: the bottom of a vertical run carries everything above it plus the accumulated weight of the cable hanging in it, while the top carries what is left after each floor has taken its share.

The reverse error costs more than money. Averaging the load along a long route makes the loaded end look acceptable, and the loaded end is where the bend into the switchroom usually sits. Work outward from the heaviest segment, note where the schedule steps down, and record the segment boundaries on the layout so the next person adding cable can see which stretch they are about to change.

The count is positions. The hardware is a separate argument

What a position count gives you is where the hangers go, not what each one is made of. That second question is settled by the deck: a beam clamp on a steel flange, a drop-in or wedge anchor in a concrete soffit, a purpose-made fixing at a bar joist panel point rather than wherever the rod happens to land, and on a post-tensioned slab nothing at all until the slab has been scanned and somebody has put permission in writing. Rod diameter, channel size and clamp working load are chosen against the load per metre from earlier multiplied by the tributary length each hanger carries, which for evenly spaced supports is simply the span.

Seismic bracing is additional to all of it and is designed rather than counted. ASCE/SEI 7 sets out how distribution systems are treated and where narrow exemptions apply on component weight and hanger length, and the local building code names the edition in force; outside the United States the equivalent obligation comes from the loading standard that code adopts. A gravity support scheme with no lateral restraint is complete for gravity and silent about everything else, which is a fine position to hold as long as it is stated on the drawing rather than assumed by whoever reads it next.

What else is warming the room, and what it is not allowed to block

The transformer is rarely alone. A UPS dissipates its own conversion losses continuously and its battery string has an ambient range of its own that is narrower than the room's. Variable-speed drives lose a published percentage of throughput each, which is small per drive and not small in a room of twelve. Panelboards, switchgear and busway all contribute; so do lighting and, in a small room, the people commissioning it. The mechanical designer needs the sum, not the transformer alone, and the method for turning equipment into sensible heat gain is standard practice set out in the ASHRAE Handbook — Fundamentals.

Whatever cools the room then has to fit around the electrical clearances rather than through them. NFPA 70 Article 110 requires working space of a stated depth, width and height in front of equipment likely to be examined or serviced live, and that space belongs to the electrician; a fan coil, a duct drop or a condensate line placed inside it is a defect regardless of how well it cools. Where the local rules are BS 7671 or AS/NZS 3000 instead, the same principle applies with different dimensions. Settle this at coordination, because the alternative is discovering it at inspection with the unit already hung.

Leave the arithmetic where the next person will find it

The load per metre each segment was designed against is worth more to the building than the tray is. Write it on the record drawing, segment by segment, with the span it was checked at and the spare capacity left over expressed in the same units. A contractor arriving in five years to add a bundle can then answer the only question that matters — how much may I add before this stops being safe — without hiring somebody to reconstruct a calculation from a cable schedule that has itself been amended twice.

Do the same with the count. Record hanger positions as installed rather than as drawn, and where a hanger moved because ductwork got there first, note the span that move created and the fact that it was checked. A run at 2.4 m centres where the layout said 3 m is not an embarrassment to bury; it is the evidence that somebody did the arithmetic when the site changed underneath them. The opposite case — a hanger quietly omitted at a services clash — leaves an effective span the tray class never covered and leaves no trace at all.

Then hand the room's heat figure to whoever maintains the building, with the transformer nameplate, the loss data it came from and the loading it was evaluated at. Load grows in switchrooms the same way it grows in tray: one addition at a time, each one small, each one signed off by somebody who could not see the last. Both numbers exist to be revisited, and neither survives being kept in one person's head.