How this instrument works
Apparent power, in kVA, is what a transformer's windings actually experience: the full current flowing through the coil, multiplied by voltage, regardless of whether that current is doing useful work. Real power, in kW, only counts the fraction of that current doing work — turning a shaft, making heat, producing light. Power factor, pf, is the ratio between the two, cos of the phase angle between voltage and current, so dividing real power by pf recovers the larger apparent-power figure the nameplate rating is built from.
The shape of the formula comes from the power triangle: three related quantities drawn as a right triangle. Real power, in kW, sits on one leg; reactive power — the current that magnetizes motor windings and transformer cores without producing work — sits on the other; the hypotenuse is apparent power, in kVA, the quantity a transformer actually has to carry. Power factor, pf, is defined as the ratio of the first leg to that hypotenuse, so solving the ratio for the hypotenuse gives kVA = kW ⁄ pf directly. Because pf is a fraction no larger than 1, dividing by it can only hold the answer steady, at pf = 1, or push it upward; it never shrinks the requirement below the kW figure.
What this number will not tell you is how much margin to leave. It is the steady-state minimum for the load exactly as measured, and says nothing about the inrush current a transformer sees the instant it energizes, which can spike to eight or ten times rated current for a few cycles, or about the extra heating that variable-frequency drives and LED ballasts cause through harmonic currents the formula does not model. Those factors call for engineering margin added on top of the bare kVA figure, not folded into it.
- Enter the Real load power (P) in kW — a wattmeter reading, a utility bill's demand figure, or the summed nameplate kW of the equipment on the circuit.
- Enter the Load power factor (pf) as a decimal between 0 and 1 — read it off a utility bill, a motor nameplate, or a clamp-meter's true-power-versus-apparent-power display.
- Read Required transformer size, kVA — this is the minimum apparent power the winding must carry for the load as entered.
- Round the result up to the nearest standard transformer rating your supplier stocks, such as 75, 100, 112.5, or 150 kVA, and add margin for growth before ordering.
Worked example — an 80 kW panel at 0.8 power factor
A distribution panel feeding a mix of induction motors and fluorescent lighting measures 80 kW of real power at a power factor of 0.8, typical for that blend of loads. The formula gives kVA = 80 ⁄ 0.8 = 100. The transformer serving this panel needs to be rated at least 100 kVA, not the 80 kVA a wattmeter alone would suggest — the extra 20 kVA is the magnetizing current the motors draw, current the copper must carry even though it produces no work at the load.
Run the same 80 kW through a transformer at a perfect unity power factor instead and the requirement drops to exactly 80 kVA, since kVA and kW converge when pf equals 1. The 25 percent gap between the two cases — 100 kVA against 80 kVA for identical useful power delivered — is the entire cost, in copper and steel, of running a facility at a mediocre power factor, which is also why many utilities meter power factor separately and add a penalty charge when it drifts low.
Questions
Why is apparent power in kVA larger than real power in kW?
Because a transformer's windings carry whatever current the load draws, whether or not that current does useful work. Real power measures work actually done — heat, light, torque. Apparent power is voltage times total current, including the reactive current that inductive loads like motors need just to build their magnetic fields. Dividing by power factor restores that reactive share, so the kVA figure is never smaller than the kW figure for the same load.
What power factor should I use if I have not measured one?
0.8 lagging is a common planning default for a mixed commercial or light-industrial load of motors, lighting ballasts, and HVAC compressors. A single induction motor running well under its rated load can be worse, down near 0.6; a purely resistive load such as electric-resistance heat sits at 1.0. Where possible, pull the actual figure from a utility bill or a clamp-meter reading rather than guessing.
Why not just add a safety margin to the kW figure instead?
Because kW and kVA describe different things a transformer must survive, and a margin tacked onto kW still ignores the reactive current in the windings. Two loads with identical real power but different power factors need genuinely different transformer sizes — dividing by pf is what converts the real-power measurement into the apparent-power figure the nameplate rating is actually defined by.
Does this result already include margin for future load growth?
No — it is the bare minimum apparent power for the load exactly as entered, with nothing held in reserve. Engineering guidance commonly adds 20 to 25 percent above this minimum for anticipated growth, plus further headroom for harmonic-heavy equipment such as variable-frequency drives, before a standard nameplate rating like 75, 100, or 150 kVA is actually selected.
Can the power factor be greater than 1?
No. Power factor is the cosine of the phase angle between voltage and current, so its magnitude is bounded between 0 and 1 — a sign of leading or lagging describes whether the load is capacitive or inductive, but only the magnitude matters for sizing. A value of exactly 1 means voltage and current are perfectly in step, and kVA equals kW.
What actually happens if the transformer is undersized?
It runs hotter than its insulation is rated for once the load approaches or exceeds nameplate kVA. Sustained current above that rating drives winding temperature past its design limit, which accelerates insulation aging and shortens the transformer's working life, sometimes sharply if the overload persists. Sizing from apparent power rather than real power alone is what keeps the winding current inside the rating.