Blog

Power Factor Correction: What Capacitor Banks Do

How a capacitor bank supplies a motor's reactive current locally, raises power factor toward 1, and removes TNB's surcharge, plus the silent failure that brings it back.

Tan Kok XinTan Kok XinElectricity Fundamentals
Open capacitor bank cabinet beside power triangles shrinking as power factor improves toward unity

Part 23 of 23 in Cobler's Electricity Fundamentals series. New here? Start with the course map.

A 75 kW chiller compressor sitting on your switchboard is drawing current it never turns into cooling. Every half of the 50 Hz cycle the motor's windings pull current in to build their magnetic field, then push it back out as the field collapses. That reactive current does no work, but it still travels the full distance from the power station, through TNB's transformers and your cables, and back again. You pay for the wire it fills, the transformer capacity it consumes, and, once your power factor slips below 0.85, a surcharge on top. Power factor correction is the fix, and a capacitor bank is how you install it.

Parts 7 and 8 of this series explained why reactive power exists and why it is not wasted energy. This is the payoff: the physical device that makes the problem go away.

What does power factor correction actually do?

A capacitor bank supplies the motor's reactive current locally, so the grid no longer has to.

A capacitor is the electrical opposite of an inductor. An induction motor's windings are inductive: the current lags the voltage, and the winding demands reactive current to magnetise. A capacitor does the reverse. Its current leads the voltage. Wire the two together on the same busbar and their reactive currents are almost exactly out of phase, half a cycle apart. When the motor's field wants current, the capacitor is discharging and hands it over. Half a cycle later the field collapses and pushes current back, and the capacitor absorbs it. The reactive energy sloshes back and forth locally between capacitor and motor at 100 Hz, twice the line frequency, instead of being hauled from the power station.

Install a bank of capacitors at your main switchboard, sized to match the reactive demand of your motors, and the reactive current stops flowing upstream. TNB's meter sees less apparent power (kVA), your power factor climbs toward 1, and the reactive-energy count that drives the surcharge falls.

How much does poor power factor cost in Malaysia?

Below PF 0.85 TNB adds a progressive surcharge to your bill, and at 0.80 it is already 7.5% of your energy, demand, and fuel-adjustment charges.

TNB requires customers supplied below 132 kV, which is nearly every commercial and industrial site at 11 kV or 33 kV, to keep power factor at 0.85 or higher. At 132 kV and above the threshold rises to 0.90. (TNB power factor page.)

Below 0.85 the surcharge is progressive. For every 0.01 you fall below the threshold, TNB adds 1.5% of your billable total, and below 0.75 each further 0.01 costs 3%. The percentage applies to your energy charge plus your demand charge plus the Automatic Fuel Adjustment. Work a real number: at power factor 0.80 you sit five steps of 0.01 below 0.85, so 5 × 1.5% = 7.5% on that total, every month, until you fix it. Our power factor surcharge article walks through more cases.

That surcharge sits on top of an already expensive demand charge. Under the RP4 tariff, medium-voltage sites pay RM89.27 or RM97.06 per kW of maximum demand each month. Poor power factor inflates the current your cables and transformer carry for the same real load, so a 1,000 kVA transformer running at 0.75 power factor delivers only 750 kW of useful work. You paid for 1,000 kVA of capacity and a quarter of the real power it could deliver is lost to carrying reactive current.

Why do motors drag power factor down in the first place?

A motor's magnetising current stays roughly constant no matter how lightly it is loaded, so an oversized motor running half-empty has terrible power factor.

At full load an induction motor runs at about 0.85 power factor. Its real power draw is high and the fixed magnetising current is a small share of the total. Unload it and the real power collapses while the magnetising current barely moves. A lightly loaded motor can sit anywhere from 0.15 to 0.40. Oversized, lightly loaded motors are the single most common reason a site's power factor is poor. So are the other reactive loads that fill a building: chillers, compressors, pumps, fans, transformers, and the ballasts in older fluorescent and HID lighting.

Why can't you just add capacitors and forget them?

Your reactive load changes through the day, and both too little and too much correction cause problems, which is why real banks switch in steps.

A factory at 7 am with one compressor running has a fraction of the reactive demand it carries at 2 pm with the whole plant online. A fixed block of capacitance sized for the afternoon would badly over-correct at dawn. Over-correction pushes power factor leading, which raises voltage on your bus, can destabilise the supply, and attracts its own utility penalty. Power factor is capped at 1.0, and the sensible target is not "maximum" but a controlled band, usually 0.95 to 0.98.

So a real installation is an automatic bank: several capacitor steps plus a power factor regulator that measures the live power factor and switches steps in and out to track the target as load changes. It keeps you comfortably above 0.85 without ever tipping leading.

What about VFDs and harmonics?

Where variable-frequency drives are present, a plain capacitor bank can resonate with the system and amplify harmonics, so you fit detuned reactors.

Part 18 covered how VFDs and other non-linear loads inject harmonic currents at multiples of 50 Hz: the 5th at 250 Hz, the 7th at 350 Hz. A capacitor bank and the inductance of your supply form a resonant circuit, and if its natural frequency lands near a harmonic already on the bus, the bank amplifies that harmonic instead of quietly correcting power factor. The result is overheated capacitors, tripped protection, and early failure.

The fix is a detuned bank: a reactor placed in series with each capacitor step, sized so the resonance sits below the lowest harmonic present. A 7% reactor tunes the bank to 189 Hz, a 14% reactor to 134 Hz, both safely under the 250 Hz 5th harmonic. Above its tuning frequency the bank behaves inductively and cannot resonate. If your site runs VFDs, and most modern plants do, detuning is not optional. Our harmonics article explains why the distortion is there to begin with.

The failure nobody notices

A capacitor bank is not maintenance-free. Individual capacitor stages age, dry out, and eventually fail, usually open-circuit and silently. A blown fuse on one step drops that step's share of reactive correction, its kVAR, out of the circuit. The regulator keeps switching the remaining steps, the bank still looks alive from the outside, and nobody on the floor sees anything. What changes is the number on next month's bill: the surcharge you eliminated two years ago quietly returns.

This is why power factor should be monitored continuously, not checked once at commissioning and forgotten. CobiNeural tracks power factor at both location and equipment level under its Energy insights, alongside maximum demand, so a decaying capacitor stage shows up as a trend the day it starts drifting, not as a surprise on a TNB bill three months later. The bank fixes the physics; monitoring keeps the fix honest.

The whole arc, in one paragraph

Twenty-three parts ago this series started with what electricity actually is, then gave it pressure and flow and a loop to run around. We separated power from energy, untangled the units they are measured in, traced why power is counted in watts, built the power triangle, and showed that reactive power is not wasted. We covered AC versus DC, met Faraday's one trick in its three costumes, the generator, the transformer and the motor, watched the war that settled the grid, and saw why it beats at 50 Hz on three phases. We followed how your meter counts and the journey from power station to plug. Then the things that go wrong: harmonics, voltage sags and swells, and the earthing and RCDs that keep faults from becoming funerals. And finally the modern toolkit: the silicon translators, batteries, and the capacitor bank you just read about. Every one of them shows up, eventually, on a TNB bill or a tripped line. Reading that bill and that plant with clear eyes is the whole point.


This is Part 23 of 23 in Cobler's Electricity Fundamentals series. Previous: How Batteries Store Energy (and Why It Is All DC).

Paying a power factor surcharge and not sure your capacitor bank is still doing its job? Book a demo and we will show you what continuous power factor monitoring looks like at location and equipment level.

FAQ

Frequently asked questions

Keep Reading

Related articles