Cable tray systems and installation field guide for electrical crews

What cable tray is and where it fits

Cable tray is a rigid metal or nonmetallic support system that carries cables and conductors along a route. It holds the cable in the open and supports its weight. It is a support system, not a raceway that encloses and protects conductors the way conduit does, and that one distinction drives almost every rule that follows.

You see tray where a lot of cables run the same path: plant process areas, electrical and mechanical rooms, commercial ceiling spaces, telecom and data center rooms. One tray run can carry power feeders, control cable, and instrumentation on the same support, and adding a circuit later is a matter of laying cable in the tray instead of pulling a fresh conduit. That accessibility is the reason the trade reaches for tray on any job where the cable count is high or the wiring is going to change.

Tray sits inside the larger wiring-method decision. The wiring-methods guide covers the full menu of raceways and cable assemblies and when each one applies. This guide is the tray half of that decision: the tray types, the materials, the load and fill rules in NEC Article 392, and the install details that make a run pass inspection and stay neat for twenty years.

What is the difference between cable tray and conduit?

The difference is support versus enclosure. Cable tray supports cables that are already rated to stand on their own, out in the open air, and lets you run dozens of them along one path. Conduit is an enclosed raceway that surrounds and protects the conductors inside it, and you usually pull a limited number of conductors through each run. Tray is a structure the cable rides on. Conduit is a pipe the wire lives in.

Tray wins on cable count, heat, and change. A single tray carries far more cable than the equivalent bundle of conduits would, the open air helps the cables shed heat, and adding or removing a circuit is fast because the tray is accessible the whole length. Conduit wins on physical protection, on running a small number of circuits a long way, and anywhere the cable needs to be fully enclosed from impact, weather, or a classified atmosphere.

The real jobs mix them. A plant runs the heavy cable on tray overhead and drops to equipment in conduit for the last few feet, where the cable needs protection and a clean termination. The honest way to choose is by cable count and change. Many cables on a shared path that will grow over the years points to tray. A handful of conductors that need enclosure and protection points to conduit. The wiring-methods guide walks the conduit families in detail.

The cable tray types

Five tray types cover most work, and they differ by how much of the bottom is open. More opening means more airflow and less material. Less opening means more support surface and more physical protection for the cable.

Ladder tray is the most common. Two side rails connected by rungs at a set spacing, usually 9 or 12 in. on center. The rungs support the cable and the open space between them lets heat off, which is why ladder is the default for power cable. Solid-bottom tray has a continuous closed bottom. It gives the most support and the most protection from falling debris, at the cost of airflow, so it suits small or sensitive cable that you want shielded. Ventilated or trough tray is the middle ground: a bottom with slots or openings that still passes air while supporting smaller cable that might sag between ladder rungs.

Wire mesh tray, the welded basket, is the fast LV and data tray. It is light, you cut and form it on site, and the open mesh keeps air moving around the cable. It is the standard for structured cabling and fiber in commercial and data center work, not for heavy power. Channel tray is a small one-piece section for a single cable or a short drop, the kind of run that does not justify a full ladder. Pick the type by what rides in it and how much air the cable needs.

Tray material: steel, aluminum, and FRP

Three materials cover the field, and you choose by environment and load. Steel is the strong, low-cost option. Galvanized steel handles indoor and ordinary outdoor work, and stainless steel goes where the chemistry is harsh and galvanizing will not last. Steel carries the most weight per dollar, which is why it shows up on heavy plant runs.

Aluminum is lighter, will not rust, and resists many atmospheres that eat steel, so it is common outdoors and in mildly corrosive areas. It is easier to handle on a long overhead run because a section weighs less, and it carries more current as a grounding path than steel does. It costs more than galvanized steel and it moves more with temperature, which matters on long runs in the sun.

Fiberglass-reinforced plastic, FRP, is the corrosive-environment tray. Where the air carries salt, acid, or chlorine, water treatment plants, coastal sites, chemical process areas, metal tray corrodes and FRP does not. FRP is nonconductive, so it cannot serve as a grounding path and you run a separate equipment grounding conductor. NEMA FG 1 covers nonmetallic tray and its mechanical, UV, and chemical properties. Match the material to the worst condition the run sees, not the average one, because the tray fails at the worst spot.

Load class and span: sizing the tray as a structure

Tray is a beam carrying a distributed load, so it is sized the way a beam is, by the load it holds and the distance between supports. NEMA VE-1 sets the manufacturing and load rules and assigns each tray a load and span class. The class pairs a support span, commonly 8, 12, 16, or 20 ft, with a working load, often around 50, 75, or 100 lb per linear ft. A wider span between supports means the tray has to carry more between bearing points, so a longer span needs a stronger class for the same cable weight.

Two numbers drive the pick: the weight of the cable fill per foot and the support spacing the building geometry allows. Estimate the cable weight from the cables you plan to install plus realistic spare for growth, then choose a class whose working load covers it at the span you can actually support. The published rating is a working load with a safety factor already built in by the standard. Do not treat it as the breaking point and do not load to it without margin.

Under-rate the tray or stretch the support span too far and you get sag, then standing water in the low spots outdoors, then bolt-hole elongation at the splices, then a run that eventually tears loose. The cheap move of one fewer support per run is the one that shows up as a sagging, ponding tray two summers later. Verify the load class and the span against NEMA VE-1 and the manufacturer's load tables for the exact tray, because the numbers vary by product.

Supports and seismic bracing

Tray hangs from the structure on supports sized and spaced for the load class. The common arrangements are the trapeze, a horizontal support hung on two rods from the structure with the tray riding on top, the cantilever bracket off a wall or column for a single-sided run, and direct wall brackets. The support spacing has to match the span the load class assumes. A tray rated for a 12 ft span that gets supported every 18 ft is no longer carrying its rated load.

Attach to real structure, not to the convenience of the moment. Tray hung off another trade's pipe hangers, off the deck flutes with the wrong fastener, or off ductwork is a callback waiting to happen. Each support has to reach back to the building steel, the concrete, or a properly anchored strut frame.

Seismic bracing is its own requirement in the jurisdictions that enforce it. A long heavy tray run is a suspended mass, and in a seismic design category that matters, it needs lateral and longitudinal bracing on a spacing the structural documents call out. That bracing is frequently value-engineered off the drawings and then caught at inspection. Confirm the seismic requirements with the structural engineer and the adopted building code, not just the electrical drawings, because the bracing for suspended systems lives on the structural side.

NEC Article 392 and what is allowed in tray

NEC Article 392 is the code that governs cable tray: where it is permitted, what wiring methods can ride in it, how full it can get, how it is supported, and how it is grounded. The first thing the article does is treat the tray as a support system for approved wiring methods, so the question is never just the tray. It is the tray plus what you are allowed to lay in it.

Tray-rated multiconductor cable is the usual occupant, cable listed and marked for cable tray use, type TC and the like, along with MC cable and other recognized assemblies. Single insulated conductors are more restricted. The article permits single conductors in tray only at 1/0 AWG and larger, and generally only in industrial establishments where qualified people maintain the system. The single-conductor rules also drive rung spacing, with a tighter spacing, commonly cited at 9 in. for 1/0 through 4/0, so the conductors are supported often enough.

The exact paragraph numbers in Article 392 shift between code cycles, and the uses-permitted list has conditions this guide does not reproduce in full. Confirm the allowed wiring methods, the single-conductor provisions, and the conditions of use against the NEC edition the jurisdiction has actually adopted and any local amendments. The cable's own listing matters as much as the article: a cable not marked for tray use does not belong in tray no matter what the article permits.

What is the fill limit for cable tray?

Cable tray fill is limited so the cables are not packed so tight they cannot shed heat. NEC 392.22 sets the limits, and they depend on the cable size and the tray type. For multiconductor power cable 4/0 AWG and smaller in ladder or ventilated tray, the common limit is 50 percent of the tray cross-sectional area from the fill table. For cable larger than 4/0, the rule shifts to a sum-of-diameters approach: the sum of the cable diameters laid in a single layer cannot exceed the tray's usable width, which keeps the large cable spread out in one layer.

Solid-bottom tray is more restrictive than ladder because it gives the cable less air, so its allowed fill runs lower for the same cable. Control and signal cable has its own, more generous allowance, commonly around 50 percent of the cross-section, since those cables carry little current and little heat. Mixed installations, where different sizes share the tray, follow the combination rules in the fill table rather than a single flat percentage.

The number people get wrong is the spare. A tray filled to the code maximum on day one has no room for the adds and changes that are the whole reason you ran tray. Size the tray with real spare capacity, often a tray one width larger than the day-one fill demands. Overfilling is also an ampacity problem, because crowded cable runs hotter, which ties straight into the ampacity rules. Confirm the specific fill percentages and the sum-of-diameters limits against NEC 392.22 for the adopted edition, and see the ampacity derating guide for what the fill does to the conductor's current rating.

Conductor ampacity in cable tray

Cables in tray follow their own ampacity rules, because how a cable sheds heat in open tray differs from how it sheds heat in a packed conduit. NEC 392.80 covers ampacity of conductors in cable tray, and the short version is that spacing and arrangement change the answer. Cables laid spread out, with space between them, carry more current than the same cables bundled tight, because the spread cable has more air around it.

The flip side is the derating you cannot dodge. Pile cable deep in a tray and the conductors in the middle run hot, so the ampacity drops just as it would in a crowded conduit. This is where overfilling bites twice: once on the fill rule and again on the current the conductor is actually allowed to carry. The detailed correction and adjustment math, the temperature corrections, and the spacing cases live in the conductor ampacity and derating guide. Run the tray ampacity there for the real numbers, and treat 392.80 plus that guide as the pair that sizes the conductor, not a habit carried over from conduit.

Can cable tray be used as a ground?

Yes, a metal cable tray can serve as the equipment grounding conductor, but only when it is listed and marked for that use and every section is bonded into a continuous path. NEC 392.60 sets the rules. The tray's side rails have to provide a minimum metal cross-sectional area for the overcurrent device protecting the largest circuit in the tray, from the table in 392.60, and that is where steel and aluminum part ways. Steel tray used as the grounding path is commonly limited to circuits with ground-fault protection at lower ampere levels, while aluminum, carrying current better, goes higher. Confirm the exact metal-area and ampere limits against 392.60 for the adopted edition.

Continuity is where it falls apart in the field. The tray is only a ground if current can travel its whole length, and every splice plate is a joint that can interrupt that path. Listed splice plates can maintain the bond, but where a joint is not bonded by the listed connection, you install a bonding jumper across it, sized per NEC 250.122 for the largest overcurrent device in the tray. Expansion splices, painted joints, and any fitting that breaks the metal-to-metal contact need a jumper.

The blunt version: a tray approved as an equipment grounding conductor with one un-bonded splice is not a ground. Fault current hits the gap and the path is gone. FRP tray is nonconductive and cannot be a ground at all, so it always carries a separate equipment grounding conductor. Many engineers run a dedicated grounding conductor in metal tray too, rather than rely on the tray, which is the conservative call on a critical system.

Splices and fittings

Tray turns and changes elevation with manufactured fittings, not field-bent rail. Horizontal and vertical elbows take the run around corners and up and down, tees and crosses branch it, reducers step the width down as the cable count drops, and risers carry it between levels. Using the manufacturer's matched fittings keeps the bend radius, the load rating, and the grounding continuity intact through the turn. A field-hacked corner loses all three.

Splice plates join straight sections and fittings. They are not just mechanical: in a tray used as a grounding path they carry fault current, so they have to be installed with the right hardware and torqued, and they are the joints that need bonding jumpers where the listed connection does not maintain the bond. A loose splice plate is both a structural weak point and a break in the ground.

Match the fitting radius to the cable. Fiber and large data cable need a generous bend radius, and a tight elbow will violate the cable's minimum bend radius even though the tray fitting fits the rail. Choose the fitting radius for the most sensitive cable in the run, then verify the cable's own minimum bend radius against its listing, since that number, not the fitting, is what protects the cable.

Thermal expansion on long runs

Metal tray grows and shrinks with temperature, and on a long run that movement is real. An aluminum tray run in the sun can move on the order of centimeters between a cold winter night and a hot summer afternoon, and aluminum moves more than steel. If the run is bolted solid end to end with no room to move, the temperature swing tears the tray off its supports, elongates the bolt holes at the splices, and buckles the rail.

The fix is expansion splice plates, slotted joints that let the sections slide as they grow. NEC 392.44 calls for expansion fittings where needed to handle thermal movement. The gap you set at each expansion joint is not arbitrary: it depends on the high and low metal temperatures the run will see and the tray temperature at the moment you install it, read off the manufacturer's gap nomograph. Set the gap for the install-day temperature so the joint has room to move both ways.

Support the expansion joint correctly. Place a support close to each side of the expansion splice, commonly within about 2 ft, and do not clamp the tray so tightly to its supports that it cannot slide. The tray has to be free to move at the joint, or the expansion splice does nothing. Long outdoor and uninsulated runs are where this gets skipped and where it shows up as damage the next season.

Installing cable in the tray

Cable goes into tray differently than it goes into conduit. You are not pulling through a pipe; you are laying cable along an open support, which is faster and gentler on the cable, but it has its own discipline. Roll or feed the cable in along the tray, support it as you go, and keep it from dragging hard over the rail edges and rungs, which can cut the jacket.

Respect the bend radius everywhere the cable changes direction, leaving the tray, entering a fitting, dropping to equipment. Power cable has a minimum bend radius that is a multiple of its diameter, and fiber and data cable need far more room than people expect. A cable kinked tighter than its rating may pass a continuity test and still fail later, and on fiber a tight bend shows up as attenuation. The cable's listing carries the real minimum bend radius; honor that number, not the radius that happens to be convenient.

Dress the cable, but do not strangle it. Use cable ties or straps to keep the run orderly and to hold cable on vertical and sloped runs, snug enough to control it and loose enough not to deform the jacket. The temptation to bundle everything into tight, tidy bunches is the wrong instinct: tight bundles trap heat and force the ampacity derating up, which defeats the open-air advantage that made tray worth using. Keep power cable spread for cooling, group like with like, and leave slack at the terminations and at every spot where the cable will move.

Keeping power and signal apart

Power and low-voltage signal cable do not always belong in the same tray. The concern is partly heat and partly interference, and partly code. Power cable running alongside unshielded control, instrumentation, or data cable can couple noise into the signal, and the higher-energy power circuit is a hazard to the low-energy one if their insulation systems are not matched.

The common solutions are separate trays or a barrier strip down the middle of one tray. Many plants run a dedicated tray for power, another for control and instrumentation, and a third for data and fiber, stacked or spaced apart, precisely so the systems stay separated and each can be worked without disturbing the others. Where they share a tray, a solid barrier keeps the power on one side and the signal on the other.

Class 2 and Class 3 circuits have a specific separation requirement from power conductors unless the low-voltage cable is rated for the power voltage present. That rule lives in the remote-control, signaling, and power-limited circuit articles, not just in 392, so a tray that looks fine under the cable tray article can still violate the separation rules for the low-voltage cable in it. The wiring-methods guide covers the location and circuit-class categories that drive these separation calls.

Tray covers and where they are required

Covers go on tray where the cable needs protection from what falls on it or shines on it. Indoors, a cover is common where the tray runs under a walkway, under another trade's leaky equipment, or anywhere debris would land on the cable. Outdoors, the driver is usually sun and weather, and the most common reason is solar.

Sunlight degrades cable jackets over time, so a tray exposed to the sky often needs a solid cover, or sunlight-resistant cable, or both. The cover keeps UV off the jacket and sheds rain and debris. The trade-off is heat: a covered tray traps more heat than an open one, and where the ampacity rules treat a covered tray more strictly, the cover that protects the cable also derates it. Vented covers are a middle path on runs where you need shade without fully boxing in the cable.

Match the cover to the reason. A short solid cover under a drip point solves one problem; a continuous cover on an exposed outdoor run solves another and adds a heat penalty you account for in the ampacity. Confirm where covers are required and how they affect ampacity against the cable listing and the adopted code, since both the protection and the derating are spec-driven.

Cable tray in data centers and industrial plants

Two settings show tray at its best, and they use it differently. In the data center, tray runs overhead in two distinct systems. Heavier ladder or solid tray carries the power feeders to the rows, and a separate wire-mesh basket grid carries the structured cabling and fiber above the cabinets. The basket is light, field-formed, and easy to add to as the room fills, which matches a space that changes constantly. The power and the data ride in separate tray for separation and so each can be worked without touching the other.

In the industrial plant, tray is the heavy hauler. Long ladder runs carry power and control cable across process areas, often hundreds of feet, with the cable count growing over the life of the plant as instruments and drives get added. The plant is where the load class, the support spacing, and the expansion splices all matter most, because the runs are long, the cable is heavy, and the temperature swings are real. It is also where single conductors in tray are permitted, since the qualified-maintenance condition the code attaches to single conductors is the plant's normal state.

Both settings share the reason tray was chosen: many cables on a shared path that keeps growing. The data center keeps power and data apart in separate basket and ladder runs; the plant keeps heavy power on tray and drops to equipment in conduit. The thread through both is accessibility, which is the whole point of running tray instead of a wall of conduit.

Firestopping tray penetrations

Where a tray passes through a fire-rated wall or floor, the penetration has to be firestopped to restore the rating the cable just punched a hole in. This is a real failure point on tray jobs because the penetration is large, the cable fill changes over time, and the firestop is easy to leave for later and then forget.

Use a tested firestop system rated for a cable tray penetration, not a tube of caulk improvised on the spot. The system is listed for a specific assembly, the wall or floor type, the tray, and the cable fill, and it has to be installed to that listing to hold the rating. The tricky part with tray is that you add and remove cable through the life of the building, so the firestop has to be a re-enterable type where future cable will pass through the same opening. A poured or sealed penetration that nobody can get back through gets hacked open by the next crew, and the rating is gone.

The blunt version: an unsealed or wrongly sealed tray penetration through a rated barrier is a code violation and a life-safety gap, and it is exactly the kind of thing an inspector walks the building looking for. Firestop the penetration to a listed system, label it, and keep it re-enterable for the adds and changes the tray exists to allow.

Why crews choose tray over conduit

The case for tray comes down to three things conduit cannot match on a high-cable job: speed, capacity, and change. Laying cable in an open tray is faster than pulling it through pipe, one tray carries the cable that would take a bank of conduits, and the run stays accessible so the next circuit goes in without new raceway.

Change is the one that pays off over years. A building wired in conduit is expensive to modify because every new circuit needs a new raceway path. A building wired on tray absorbs adds and moves almost for free, since the support is already there with spare capacity, if you sized it with spare. That is why plants, labs, and data centers, the places that rewire themselves constantly, run tray as the default and reserve conduit for the protected drops.

Tray is not the answer everywhere. It needs the cable to be rated for open support, it needs more vertical space than a conduit bank, and it gives less physical protection. On a job with a few circuits, no growth, and a need for enclosure, conduit is the right call. The advantage is real where the cable count is high and the wiring will move, and it disappears where it will not.

What does an inspector check on cable tray?

An inspector walks a tray run looking at the same short list every time, and a crew that knows the list builds to pass it. The first checks are fill and support: is the tray overfilled past the 392.22 limit, and is it supported at the spacing its load class requires, with no sag or ponding. Both are visible from the floor and both are common write-ups.

Grounding and bonding come next on a tray used as an equipment grounding conductor. The inspector looks for the listing mark, for splice plates installed and tight, and for bonding jumpers across the joints that need them, especially at expansion splices and any non-bonded fitting. A tray claimed as a ground with a missing jumper is a fail.

Then the details that separate a clean job from a marginal one: expansion splices present and set on long runs, covers where the spec or exposure requires them, power and signal separated where the rules demand it, cable bend radius respected at the drops, and firestop installed and listed at every rated penetration. The cable listing matters too, since a cable not marked for tray use in the tray is a finding regardless of how neat the run looks. Build the run so each of these is right the first time and the inspection is a walk-through, not a punch list.

Workmanship: the run that passes on looks

A tray run that looks right usually is right, because the same care that makes it neat makes it correct. Straight runs that are actually straight, fittings instead of field-bent corners, splices tight and aligned, supports plumb and evenly spaced. An inspector who sees a sloppy run starts looking harder at everything else, and an owner who sees a clean run trusts the parts they cannot see.

Cable dressing is the visible craft. Cable laid in flat, combed out at the turns, tied at a regular spacing, with power grouped away from signal and slack left where it is needed. On a basket tray full of data and fiber, the dressing is the job: a combed, labeled, color-grouped basket is the difference between a room that can be worked in five years and a rat's nest nobody will touch.

Label the cable and the circuits. A tray carries many circuits on one path, and without labels the next person cannot tell which cable is which without tracing it end to end. Label at the terminations, at the pull points, and where cable leaves the tray, so the run can be worked without a guessing game. The labeling is not decoration; it is what makes the accessibility that justified the tray actually usable later.

What to document

A tray run is a shared, long-lived asset that other people will modify for years, so the record is what lets them do it without re-engineering the whole run. Capture the tray type and material, the load and span class, the support spacing, the day-one fill and the spare capacity, the grounding method, and where the expansion splices and firestops are.

The fill and spare are the entries people skip and later need most. The next crew adding a circuit has to know how much room is left before they overfill the tray or push the ampacity past its limit, and that is impossible to judge by eye on a loaded run. Record what is in the tray and what the tray is rated to hold, so the add-a-circuit decision is a lookup, not a recalculation from scratch.

Common mistakes

• Filling the tray to the 392.22 maximum with no spare, so the next add overfills it and pushes ampacity past its limit.
• Claiming the tray as an equipment grounding conductor but leaving splices and expansion joints without bonding jumpers, so the ground path is open.
• Under-rating the load class or stretching the support span past what the class allows, which leads to sag, ponding, and bolt-hole elongation.
• Running long metal tray with no expansion splices, so summer-to-winter movement tears it off the supports.
• Mixing power and signal in one tray with no barrier or separation where the circuit class requires it.
• Forcing cable past its minimum bend radius at the drops and fittings, which damages power cable and attenuates fiber.
• Leaving rated wall and floor penetrations without a listed, re-enterable firestop system.
• Putting cable not marked for cable tray use into the tray, regardless of how neat the run looks.

Field checklist

Standards and references

NEC Article 392, in NFPA 70, is the governing code for cable tray: the uses permitted, the wiring methods allowed in tray, the fill in 392.22, the supports and expansion fittings in the installation provisions, the ampacity in 392.80, and the grounding and bonding in 392.60. The exact paragraph numbers move between code cycles, so confirm them against the edition the jurisdiction has adopted and any local amendments before citing them on a submittal.

NEMA VE-1 is the manufacturing and load standard, where the tray load and span classes come from, and NEMA VE-2 is the installation guidelines, the field companion that covers handling, support, and the expansion-gap nomograph. NEMA FG 1 covers nonmetallic FRP tray for corrosive environments. For ampacity, pair 392.80 with the conductor ampacity and derating guide, since the spacing and fill drive the current rating. Bonding jumpers across tray joints are sized per NEC 250.122 to the largest overcurrent device in the tray.

The cable's own listing controls as much as the article: a cable has to be marked for cable tray use to belong in tray, and its minimum bend radius comes from that listing. Manufacturer load tables, fitting radii, and expansion-gap nomographs are product-specific, so verify against the tray you are actually installing. Where a project specification is tighter than the code minimum, the specification governs.

Units, terms, and conversions

Cable tray spans a few naming conventions across the load tables, the code, and the manufacturer sheets, so the same idea reads differently depending on the document in front of you.

Load class is given as a working load in pounds per linear ft paired with a support span in ft in NEMA VE-1, with metric sheets using kilograms per meter and meters. Fill is a percentage of the tray cross-sectional area for smaller cable and a sum of cable diameters for larger cable. Tray width and depth are in inches in US sheets and millimeters in metric ones. Rung spacing, the distance between ladder rungs, is given in inches, commonly 9 or 12 in. on center.

Cable tray

A rigid support system that carries cables and conductors along a route in the open, governed by NEC Article 392

Ladder tray

Two side rails joined by rungs; the most common type, with the best airflow for power cable

Wire mesh / basket tray

Welded mesh tray, light and field-formed, the standard for LV, data, and fiber

Load and span class

NEMA VE-1 rating pairing a working load (lb per linear ft) with a support span (ft)

Tray fill

The portion of the tray cross-section the cable may occupy, limited by NEC 392.22

EGC

Equipment grounding conductor; a listed metal tray can serve as one when every joint is bonded

Expansion splice plate

A slotted joint that lets a tray run slide as it grows and shrinks with temperature

FRP

Fiberglass-reinforced plastic tray for corrosive environments; nonconductive, needs a separate EGC

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is the difference between EMT and rigid conduit?
• What is conductor ampacity?
• Why is cable management important in a data center?
• What is busway?
• What is a conduit body?