How this instrument works
Power factor is the ratio of real power — the kilowatts that actually turn a shaft or light a lamp — to apparent power, the kVA a utility must generate to make that real power available. The gap between them is reactive power, measured in kVAR, and it exists because motors, transformers, and fluorescent ballasts store energy in magnetic fields every half-cycle and hand it back the next. Because PF equals the cosine of the phase angle φ between voltage and current, tan φ — the tangent of that same angle — is exactly the ratio of kVAR to kW. That is why the formula subtracts two tangents: tan(cos⁻¹PF₁) is how many kVAR ride along with each kW at the present factor, and tan(cos⁻¹PF₂) is how many would ride along at the target. Multiply the load into that difference and the result is the kVAR a capacitor bank must supply to close the gap.
Capacitors correct a lagging factor because they draw current that leads voltage by 90°, the mirror image of an induction motor's magnetizing current. Installed at the load, a bank supplies the reactive current locally instead of asking the utility's wires and transformers to carry it from the substation, which is precisely why utilities meter this ratio and penalize industrial accounts running much below about 0.90 — a low-PF customer draws more amperes per billable kilowatt, forcing the utility to size conductors for current that does no useful work. The second stage, C = Qc ⁄ (2πfV²), turns the kVAR requirement into a physical part: a capacitance sized against the system's own frequency and voltage, because the reactive output of a fixed capacitor rises with the square of the voltage across it.
The capacitance returned here is a single-phase equivalent — what one capacitor would need if it alone supplied all the kVAR across the full line voltage. A real three-phase bank normally splits that total across three units, wired wye or delta, and delta-connected capacitors see line voltage rather than phase voltage, which changes each unit's required µF even though the bank's total kVAR stays the same. Push the target above roughly 0.98 and correction can overshoot into a leading factor at light load, raising bus voltage rather than lowering losses — the common mistake is sizing for the worst-case low hour without checking the lightest-load hour, and on feeders carrying variable-frequency drives a fixed bank can also tune into resonance with harmonics already on the line.
- Enter the load's Real power in kW — the wattmeter or nameplate kW reading, not the apparent kVA off the utility bill.
- Enter Current power factor as measured or read from the utility statement, between 0 and 1 (0.75 for a plant heavy with lightly loaded motors).
- Enter Target power factor — most tariffs stop penalizing between 0.90 and 0.95, which is the usual target rather than 1.00.
- Enter System voltage and System frequency exactly as they appear on the switchgear nameplate; both feed directly into the capacitance step.
- Read Reactive power correction needed, kVAR for the bank rating, and Required capacitance (single-phase equivalent) for the µF to specify per unit.
Worked example — correcting a 50 kW load from 0.75 to 0.95 PF
A machine shop's main feeder reads 50 kW of real power (realPower = 50000 W) at a stubborn 0.75 power factor, mostly from lightly loaded induction motors on the lathes and mills. The utility tariff stops penalizing at 0.95, so the target power factor is set to 0.95. At 0.75 PF, tan(cos⁻¹0.75) works out to 0.8819 kVAR for every kW; at 0.95 PF that drops to 0.3287 kVAR per kW. The difference, 0.5532, multiplied by the 50 kW load, gives Reactive power correction needed, kVAR = 27.6616499255 — call it 27.66 kVAR of capacitive compensation the bank must supply.
The shop runs a 480 V, 60 Hz service, so System voltage = 480 and System frequency = 60 feed the second formula, C = Qc ⁄ (2πfV²). Plugging in 27.6616499255 kVAR at 480 V and 60 Hz returns Required capacitance (single-phase equivalent) = 318.467037016 µF. In practice an electrician would not buy one 318 µF unit; a three-phase bank split into three units of roughly 106 µF each, wired delta and rated for at least 480 V, delivers the same total 27.66 kVAR and brings the plant's billed power factor up to 0.95.
Questions
Why does the formula use the tangent of an inverse cosine instead of the power factor directly?
Because power factor is a cosine, not a ratio of reactive to real power — the ratio actually needed is the tangent of that same phase angle. Taking cos⁻¹PF recovers the angle φ, and tan φ converts it into kVAR per kW. Subtracting the target's tangent from the current one and multiplying by real power gives the reactive power a capacitor bank must add, in one line, without computing kVA or the angle separately.
Does correcting all the way to 1.00 power factor save much more than stopping at 0.95?
Only marginally, and rarely enough to pay for itself. Between 0.75 and 0.95, tan(cos⁻¹PF) falls from 0.882 to 0.329, a drop of 0.553. Pushing from 0.95 to 1.00 removes only another 0.329, a smaller gain from a similarly sized bank, and most tariffs stop billing a penalty once an account clears roughly 0.90 to 0.95, so the last few percent buys little.
What happens if the capacitor bank ends up oversized for the load?
The plant's power factor can swing leading rather than settling near unity, especially overnight or on weekends when real load — and the inductive reactance it carries — drops while a fixed bank keeps supplying the same kVAR. A leading factor raises bus voltage instead of trimming losses, and on feeders with variable-frequency drives or other harmonic-producing loads it can also excite resonance with the system's own inductance.
Why does the required capacitance depend on the square of the system voltage?
Because the reactive power a capacitor produces is Qc = V²ωC — it rises with the square of the voltage across its plates, not linearly with it. Rearranged to solve for C, that squared term lands in the denominator, so doubling the system voltage needs only a quarter of the capacitance to deliver the same kVAR. It also means a bank rated for one voltage cannot be relabeled for another without recalculating its µF.
Is the capacitance this calculator returns what gets ordered off the shelf?
It is a single-phase equivalent — the µF one capacitor would need if it alone supplied all the kVAR at full system voltage. Real installations split that load across three units in a wye or delta bank, and delta-connected capacitors see line-to-line voltage rather than phase voltage, so the per-unit rating an electrician specifies is derived from this figure rather than copied straight from it.
Does correcting power factor on residential equipment make economic sense?
The arithmetic works the same for any single-phase load, but this instrument targets three-phase industrial and commercial service, where utilities meter and bill reactive power. Residential meters almost never charge for it, so correcting the power factor on, say, a well pump has no tariff to offset — the calculation is identical, but the economic case behind it usually is not.