Running Romex (nonmetallic sheathed cable) through conduit seems straightforward until the conduit fill rules enter: the electrical code limits how much of a conduit's cross-section conductors may occupy, and Romex, being a flat-ish oval cable rather than round conductors, complicates the calculation. In commercial work, where inspectors enforce the rules strictly and the installations must last, understanding conduit fill capacity ratios for sheathed cable is not optional trivia; it is the difference between a compliant installation and a failed inspection, or worse, an overheated raceway.
Why Fill Limits Exist
Conduit fill limits serve two purposes: heat dissipation and physical protection. Conductors generate heat under load, and a conduit packed tight traps that heat; the fill limits keep enough air space for heat to dissipate so the conductors stay within their temperature ratings. The limits also prevent mechanical damage during pulling: overfilled conduit means excessive pulling tension, which stretches and damages conductor insulation, creating faults that appear months or years later.
The code expresses fill as a percentage of the conduit's internal cross-sectional area: roughly 53 percent for a single conductor, 31 percent for two, and 40 percent for three or more. These percentages look arbitrary but encode the geometry of packing circles in a circle with room for heat and pulling; they are the product of extensive testing and field experience.
Romex as a Single Conductor for Fill
The key rule for sheathed cable in conduit: a cable assembly counts as a single conductor, and its cross-sectional area is computed from its major diameter (the widest dimension of the oval). This is conservative by design: the oval cable in a round conduit wastes space, and the code accounts for it by treating the cable as the circle its widest dimension describes. Measure the cable's actual major diameter; nominal sizes understate it enough to matter.
The practical consequence is that Romex fills conduit fast. A cable that looks like it slides in easily may already exceed the fill percentage, because the calculation uses the major-diameter circle, not the cable's actual oval area. Always calculate; never eyeball.
Working the Calculation
The procedure: find the conduit's internal area from the conduit tables, compute the cable's equivalent area from its major diameter (π × (d/2)²), divide cable area by conduit area, and compare against the applicable fill percentage. For multiple cables in one conduit, sum their equivalent areas and use the multi-conductor percentage. If the result exceeds the limit, upsize the conduit or reduce the cable count.
Note the common commercial reality: Romex is generally limited to specific uses and locations by code, and many commercial installations use individual conductors in conduit rather than sheathed cable. Where Romex does appear in commercial work (short protective sleeves, specific permitted applications), the fill calculation still applies in full; the inspector will check it.
- Measure actual cable diameters; manufacturing variation and jacket thickness differ from nominal.
- Count all cables in the conduit, including grounds and spares; everything occupies space.
- Apply derating for multiple current-carrying conductors separately; fill and derating are independent requirements.
- Document the calculation; inspectors respect shown work and question round numbers.
Sleeves Versus Complete Conduit Systems
Much Romex-in-conduit in practice is sleeving: short conduit sections protecting the cable where it passes through masonry, emerges from the ground, or needs physical protection. Short sleeves have practical considerations beyond the fill tables: the cable must still pull through without damage, and very short sleeves are more forgiving than long runs. But the code's fill limits apply regardless of length; a sleeve is still a raceway, and the inspector's tape measure does not care about your intentions.
For long conduit runs with sheathed cable, reconsider the wiring method: individual conductors (THHN and similar) in conduit pack more efficiently, dissipate heat better, and pull easier than an equivalent Romex assembly. The fill math usually shows why the pros wire commercial conduit with singles.
Safety and Inspection Realities
Overfilled conduit is a genuine fire hazard, not a paperwork violation: trapped heat degrades insulation over time, and damaged insulation from forceful pulling creates arcing faults. The fill limits are written in the language of percentages but enforced in the currency of safety. In commercial occupancies with higher loads and stricter inspection, the margins matter more, not less.
Work de-energized whenever possible, verify with a tester before touching, and follow lockout procedures on commercial panels. The conduit fill calculation happens at the desk; the safe work practices happen everywhere else. Both are part of the same professional standard.
Count the cable as one conductor at its major diameter, calculate the percentage honestly, upsize the conduit when the math says so, and prefer individual conductors for real conduit runs. Fill ratios are the code's way of keeping heat and mechanical stress within what the materials tolerate; respect them and the installation lasts.
Related Guides
- Sizing Neutral Wires for Non-Linear Loads: In a balanced three-phase system, the neutral carries only the imbalance between phases, which is why neutrals were traditionally sized smaller than
- Three-Wire Cable Sizing for Multiway Switched Circuits: Multiway switching, controlling one light from two or more locations, runs on three-wire cable between the switches: two travelers plus a common
- Derating Conductor Ampacity in Multi-Cable Bundles: A conductor's ampacity, the current it can carry safely, is really a temperature rating in disguise. The copper can handle far more current than the
- Sizing Branch Circuit Breakers for Continuous Loads: A continuous load, one that runs for three hours or more at a stretch, heats a breaker differently than an intermittent one, and the breaker must be