Energy Academy
AC Fundamentals3 / 10

Power Factor & Reactive Power

Real vs reactive power, why power factor matters, and how to improve it.

10 min read · Jacob Willis, Net Zero Lead · Last reviewed July 2026


On an AC system, a site can draw more current than the useful work it performs can explain. The extra current is not stolen and not wasted as heat on site; it flows back and forth doing something necessary but unproductive, and the network still has to carry it. Power factor is the number that measures how much of what you draw is doing real work, and because networks charge for carrying the unproductive part, it is one of the few electrical concepts with its own line on the bill.

Real, reactive and apparent power

Motors, transformers and anything with a coil need a magnetic field to operate. Building and collapsing that field 50 times a second takes current that surges into the device and back out again, transferring no net energy. This gives AC systems three related quantities:

  • Real power (kW): the power doing actual work: turning shafts, making light and heat. This is what the first lesson called power.
  • Reactive power (kVAr): the back-and-forth flow sustaining magnetic fields. Necessary for the equipment, useless as output.
  • Apparent power (kVA): the total the network must be sized to carry, combining both.

The classic analogy is a glass of beer. The liquid is real power, the froth is reactive power, and the glass must be big enough for both. Power factor is the ratio of liquid to glass:

Power factor = real power (kW) ÷ apparent power (kVA)

A power factor of 1.0 means every amp does useful work. Typical uncorrected industrial sites sit between 0.7 and 0.9, dragged down mainly by lightly loaded motors.

Worked example — what a poor power factor costs in capacity
Given
  • A site's real load: 400 kW
  • Its power factor: 0.80
  • Its supply is charged and sized in kVA
Find
The apparent power drawn, and what improving to 0.95 would free up.

Drag the power factor below and watch the triangle change shape: the useful work (kW) stays fixed while the reactive and apparent power grow as the power factor falls.

The power triangle, live: drag the power factor
Apparent 500 kVA: what the supply must carryReal 400 kW: the useful work (fixed)Reactive300 kVArφ · pf = cos φ = 0.80
Apparent power
500 kVA
Reactive power
300 kVAr
Freed by correcting to 0.95
79 kVA

A site doing 400kW of useful work. The supply, cables and transformer must all carry the kVA, and larger tariffs charge for it, which is why correcting a poor power factor (the lesson's 0.80 → 0.95 example frees ~79 kVA) buys back capacity without saving a single kWh.

Why it costs money

Poor power factor is charged in up to three ways, depending on the tariff. Larger sites pay for their agreed supply capacity in kVA, so froth occupies capacity you are paying to reserve. Many tariffs add explicit reactive power charges once the kVArh drawn passes a threshold (commonly when power factor drops below about 0.95). And the extra current causes real, if modest, additional heating losses in the site's own cables and transformers. There is also a hard limit: a site near its agreed capacity may be refused new connections, and correction is often far cheaper than a supply upgrade.

How correction works

The fix is elegantly passive. Capacitors draw reactive power in the opposite sense to motors and coils, so a capacitor bank sized to the site's reactive load cancels most of the froth locally. The reactive current then circulates between the capacitors and the motors instead of flowing through the meter and up the network. Automatic banks switch capacitance in steps as load changes, holding the site near a target power factor. Correction equipment is a mature, low-maintenance technology, and where reactive charges or kVA headroom are real problems, paybacks of one to three years are common; the site's actual half-hourly data and tariff make that case precisely, as the demand lesson later in this course shows.

Power factor saves charges, not energy

Be precise about what correction buys, because overstating it undermines credibility. It reduces kVA demand, reactive charges and cable losses; it does not meaningfully reduce the kWh your equipment consumes to do its work. The kettle boils no faster. Present it as a network-cost and capacity measure, sized from the bill, and the case stands on solid ground.

With single-phase AC understood, the next module scales up to how real sites are actually supplied: three phases at once, and why that arrangement won.

Sources and further reading