Poor power factor costs money, and a capacitor bank is the most common hardware used to fix it. Motors and other inductive loads draw reactive power that does no useful work but still loads the wires and the utility's system, and that reactive burden is what drags power factor down. A capacitor bank supplies that reactive power locally instead, so the utility does not have to. This guide explains how fixed and switched kVAR banks correct power factor, how an automatic controller decides how much to switch in, and the resonance and detuning cautions that separate a good installation from a troublesome one.
PF Correction Capacitor Bank in one line: A power factor correction capacitor bank is an assembly of capacitors that supplies reactive power (measured in kilovar or kVAR) locally to offset the reactive demand of inductive loads such as motors. By providing that reactive power near where it is consumed, the bank raises the facility's power factor, reduces current in the supply, and avoids utility power-factor penalties. Banks are either fixed or switched in steps by an automatic controller, and on harmonic-rich sites they usually include detuning reactors to avoid resonance.
Inductive loads such as motors draw a lagging reactive current that the source must supply even though it does no real work. A capacitor is the electrical opposite of an inductor: it draws a leading reactive current. Place capacitors near the inductive load and the two reactive currents largely cancel, so the reactive power circulates locally between the capacitor and the motor instead of being drawn all the way from the utility. The bank is sized in kVAR to supply enough leading reactive power to lift the power factor from its uncorrected value up to a target, commonly around 0.95.
The most straightforward arrangement is a fixed capacitor bank, a set of capacitors permanently connected, often right at a large motor or at a motor control center. Fixed banks are cheap and simple and work well where the reactive demand is roughly constant. Their weakness is that they keep supplying the same kVAR regardless of load, so if the load drops away the fixed capacitors can over-correct and push the power factor leading, which some utilities penalize and which can raise voltage undesirably.
Where load varies, an automatic or switched capacitor bank is used. It divides the total capacitance into steps, each switched in or out by a contactor or thyristor under the command of a power-factor controller that measures the site's actual power factor and adds or removes steps to hold it near the target. This tracks changing load without over-correcting at light load, at the cost of more hardware, control complexity, and switching duty on the capacitors and their switching devices.
The catch with capacitor banks is that a capacitor and the inductance of the upstream system form a resonant circuit, and if that resonant frequency lands near a harmonic frequency present on the system, the harmonic can be dramatically amplified. On a site with variable frequency drives, rectifiers, or other harmonic-producing loads, a plain capacitor bank can excite a resonance that magnifies harmonic voltages and currents, overheating the capacitors, blowing fuses, and distorting the supply far worse than the low power factor it was meant to cure.
The standard defense is a detuned capacitor bank, in which a small reactor is placed in series with each capacitor step to shift the resonant frequency below the lowest significant harmonic, usually below the fifth harmonic. The detuning reactor tames the resonance and also limits the inrush when a step switches in. On harmonic-heavy installations, detuned banks are the norm rather than the exception, and specifying a plain bank there is a classic and expensive mistake.
Beyond resonance, capacitor banks demand attention to switching transients, discharge time between switching operations, ambient temperature and ventilation, and the interaction with any generators or standby supplies. Capacitors must be allowed to discharge before being re-energized, and a bank left connected to a de-energized motor can self-excite the motor as it coasts down. These are the details that make a capacitor bank a piece of engineered equipment rather than a simple bolt-on, and getting them wrong turns a cost-saving device into a source of nuisance trips and equipment stress.
A capacitor bank only saves money while it is actually working, and banks fail quietly. Individual capacitor cans degrade or blow their internal fuses, contactors on switched steps wear out, and detuning reactors and controllers can drift, all without stopping the process. The first sign is often a power-factor penalty reappearing on the utility bill weeks after the correction silently stopped, long after anyone could have caught it on site.
This makes power-factor correction a natural thing to monitor rather than to install and forget. A power-factor controller or a metering device at the service typically exposes the measured power factor, the reactive power, and the number of steps in service over a communications interface, and those values tell you at a glance whether the bank is doing its job. A step that stops switching, a power factor that never reaches target, or capacitor current that has dropped are all readable symptoms of a failing bank.
A cloud SCADA platform such as Merobix can pull those power-factor, kVAR, and step-status tags from meters and controllers across every facility and trend them alongside the loads that create the reactive demand. That lets an operator see correction degrade in real time, catch a failed capacitor step before the next penalty appears on the bill, and confirm that a bank is not over-correcting into a leading power factor at light load, turning what is usually a set-and-forget device into something with the same visibility as any other critical asset.
Inductive loads such as motors draw a lagging reactive current that does no useful work but still loads the supply. A capacitor draws a leading reactive current that is the opposite, so placing capacitors near the load lets the two reactive currents cancel and keeps the reactive power circulating locally instead of being drawn from the utility. This raises the power factor, lowers the current in the supply, and avoids utility power-factor penalties.
A fixed bank is permanently connected and always supplies the same reactive power, which suits loads that stay roughly constant but can over-correct into a leading power factor when the load drops. A switched or automatic bank divides the capacitance into steps that a power-factor controller adds or removes to track the changing load and hold the power factor near a target. Switched banks handle varying loads better at the cost of more hardware and control complexity.
A capacitor and the system inductance form a resonant circuit, and if that resonance lands near a harmonic frequency present on a site with drives or rectifiers, the harmonic can be dramatically amplified, overheating capacitors and distorting the supply. A detuning reactor placed in series with each capacitor step shifts the resonance below the lowest significant harmonic, usually below the fifth, which prevents the amplification and also limits switching inrush. On harmonic-heavy sites detuned banks are standard.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.