Power Factor Calculator

Only the parts of the bill that are billed on kVA or carry a power factor adjustment — the kilowatt-hour line does not move at all. The energy meter registers real power, and capacitors change reactive power, not real power. What they do change is current, and therefore the demand a kVA meter records and the I²R heating in the conductors on your side of the meter. On a rate schedule with no reactive term whatsoever, the entire return is that loss reduction inside your own building, and on a short feeder run it is usually far too small to repay a bank.

Yes, and past unity the current starts climbing again — leading this time instead of lagging. A demand meter that bills in kVA charges a leading 0.90 exactly as it charges a lagging 0.90, so the money runs back out the other side. The more expensive failure is self-excitation: a capacitor left connected to an induction motor that has just been switched off feeds it magnetizing current while it coasts, and the motor acts briefly as a generator, producing voltages high enough to damage the winding and to reclose onto out-of-phase. That is why an individual capacitor is switched by the motor's own contactor and sized below the motor's no-load magnetizing current, not at the motor's full-load kVAR.

Because that 0.98 is displacement power factor and the figure your meter is unhappy about is true power factor. A six-pulse drive front end draws current in narrow pulses twice a cycle: the fundamental component sits almost in phase with the voltage, which is what makes the displacement number look excellent, while the harmonic current in those pulses carries no real power and drags the true power factor down to something far lower. Capacitance corrects only the displacement part, and it simultaneously offers a low-impedance path to the harmonics — the bank and the supply transformer's leakage reactance form a parallel resonant circuit, and if that resonance lands near the 5th or 7th harmonic the capacitor current can rise until fuses clear. A harmonic study, and usually a detuned reactor in series with the bank, is what that plant needs rather than plain capacitance.

At the motor if you want the current reduction in the feeder that serves it; at the switchboard if all you want is the utility's meter to read better. An individual capacitor at a motor unloads every conductor, contactor and transformer between it and the service, and it switches with the motor so it can never be left connected to a stopped load. A central automatic bank with contactor-switched stages follows a plant load that varies through the day and avoids the self-excitation problem entirely, at the cost of leaving every feeder downstream of it carrying the same current it always did.

Because each capacitor stands across a different voltage, and the reactive power it produces goes as the square of that voltage. In a delta bank every unit sees the full line-to-line voltage; in a wye bank each sees that voltage divided by √3, so a wye bank needs three times the capacitance to make the same kVAR. This is exactly why capacitors are ordered in kVAR at a stated voltage rather than in microfarads: the kVAR is a property of the bank, the microfarads are a property of the connection, and a supplier asked for microfarads has to guess which one you meant.

Both, usually — they are measuring different things at different places over different windows. A utility power factor is normally derived from the month's registered kVARh against kWh, or from the power factor during the single interval that set your billing demand, so a plant that runs badly for one shift and well for two can bill at a figure it never actually shows on a panel meter. The metering point differs too: the utility measures on the primary side, which includes the magnetizing current of your own service transformer, and that current is reactive, constant, and entirely absent from the reading you take downstream of it.