Power Factor Explained

Real power P, measured in watts, represents average energy transfer that produces work, heat, light, or another net conversion. Apparent power S, measured in volt-amperes, is the product of RMS voltage and RMS current for single-phase circuits. It describes electrical loading on conductors and equipment even when not all of that product becomes average real power.

Reactive power Q, measured in vars, describes alternating energy exchange associated with electric and magnetic fields under sinusoidal conditions. Inductors store magnetic energy and capacitors store electric energy, returning it during another part of the cycle. Ideal reactive elements consume no average real power, yet their current contributes to conductor heating, voltage drop, transformer loading, and source capacity.

For balanced sinusoidal systems, the power triangle satisfies S^2 = P^2 + Q^2. The triangle is a representation of complex power, not a physical flow diagram. Sign conventions for leading and lagging reactive power vary, so state whether inductive Q is treated as positive and how exported real power is represented.

For a resistive sinusoidal load, current is in phase with voltage. Their instantaneous product remains nonnegative over a cycle, and ideal power factor is one. For an inductive load, current lags voltage; for a capacitive load, current leads. Part of the instantaneous power then reverses as stored field energy returns to the source.

Displacement power factor is cos phi. At phi = 0 degrees it is one, at 60 degrees it is 0.5, and at 90 degrees an ideal reactive element transfers zero average real power. The sign of phi or Q identifies leading versus lagging, while the magnitude of PF alone does not. Reporting 0.8 without the direction omits information needed for correction.

Real devices combine resistance and reactance. Motors, transformers, ballasts, and lightly loaded magnetic equipment commonly draw lagging current. Overexcited synchronous machines and capacitor banks can supply leading reactive power. Network impedance and voltage level influence the operating point, so a nameplate or one-time value may not represent all conditions.

Nonlinear loads draw current that is not sinusoidal even when supply voltage is nearly sinusoidal. Rectifier-capacitor inputs, variable-speed drives, switched-mode supplies, and electronic lighting can draw pulsed or harmonically rich current. The fundamental current may have a small phase displacement while total RMS current is much larger because of harmonics.

True power factor remains P divided by total RMS volt-amperes. It reflects both displacement and distortion. Under suitable voltage conditions it can be understood as displacement factor multiplied by a distortion factor, but practical measurement should follow applicable power-quality definitions. A simple phase-angle meter can overstate PF when current distortion is significant.

Capacitors can correct fundamental reactive power but do not automatically remove harmonic current. They may interact with system inductance and create resonance, raising voltage or current at certain frequencies. Nonlinear installations can require harmonic studies, detuned banks, filters, active correction, or equipment-specific solutions rather than a basic kvar calculation alone.

For the same real power and voltage, lower power factor requires greater RMS current. In a single-phase sinusoidal approximation, I = P/(V PF). In balanced three-phase systems, I = P/(sqrt(3) V_LL PF), using line-to-line voltage. Higher current increases I-squared-R losses and voltage drop and consumes conductor, transformer, generator, switchgear, and utility capacity.

Correcting power factor can reduce current upstream of the correction point, release capacity, and improve voltage. It does not reduce the real power required by an unchanged load, and it may not reduce current in conductors between the load and a remotely located capacitor. Placement matters. Utility billing rules also vary by tariff and jurisdiction; technical improvement does not guarantee a specific financial outcome.

Very light loading can lower motor or transformer power factor while real efficiency follows a different curve. Power factor and efficiency are not synonyms. Efficiency compares useful output with real-power input; power factor compares real power with apparent power. A device can have high efficiency and modest PF, or high PF and poor efficiency.

Related in this workflow: Power Factor Calculator, Electrical Power Calculator.

For a sinusoidal lagging load, required capacitor reactive power can be estimated as Q_c = P(tan phi_1 - tan phi_2), where phi_1 and phi_2 correspond to initial and target displacement factors. The capacitor supplies leading reactive power locally, reducing net reactive demand seen upstream. The calculation requires consistent real power and operating condition.

A target of exactly unity is often unsuitable because load varies and fixed capacitance can produce leading PF at light load. Automatic stepped banks track varying demand, while individual motor capacitors require coordination with switching and self-excitation risks. Voltage, frequency, duty, temperature, tolerances, discharge, protection, and switching transients affect equipment selection.

Correction at low voltage, medium voltage, individual loads, groups, or a main bus has different effects on losses and capacity. A one-line diagram and load profile are needed to choose location. In harmonic environments, calculate resonance and capacitor current stress. Applying a capacitor solely from one PF reading can create overvoltage or equipment damage.

Consider a balanced three-phase load using 100 kW at 400 V line-to-line with lagging power factor 0.75. Apparent power is 100/0.75 = 133.3 kVA. Approximate line current is 100,000/(sqrt(3) x 400 x 0.75) = 192.5 A, neglecting waveform distortion and voltage imbalance.

The initial angle is acos(0.75), and tan phi_1 is about 0.882. For target PF 0.95, tan phi_2 is about 0.329. Required fundamental compensation is 100 x (0.882 - 0.329) = 55.3 kvar. At the target, apparent power is 105.3 kVA and line current falls to about 152.0 A at the same voltage and real power.

The example does not select a 55.3 kvar bank directly. Practical steps, voltage tolerance, harmonic derating, switching, minimum load, location, and utility requirements must be reviewed. Measurements over representative operating periods should confirm that 100 kW and 0.75 PF are not transient or distorted readings.

A true-RMS power analyzer computes real power from instantaneous voltage and current and apparent power from RMS quantities over a defined interval. Confirm wiring configuration, phase sequence, current-transformer ratio and orientation, voltage range, frequency, bandwidth, and whether the instrument reports displacement or total PF. Incorrect CT polarity can reverse power or reactive signs.

Three-phase systems may be unbalanced, so one phase is not necessarily representative. Measure all required channels with a suitable connection method. Distorted or rapidly varying loads need adequate sampling and aggregation. Demand-interval PF, instantaneous PF, fundamental PF, and utility billing PF may differ.

Measurement uncertainty matters near tariff or protection thresholds. Sensors have ratio and phase errors, and low current can reduce accuracy. Record operating state, load schedule, voltage, harmonic distortion, instrument class, and aggregation interval. A single unlabeled display value is weak evidence for system modification.

Do not use PF = cos phi when waveforms are significantly distorted without clarifying that it is displacement factor. Do not confuse kvar with kW or kVA. Do not omit leading or lagging direction. Do not assume a capacitor improves real efficiency or removes harmonics. These mistakes arise from reducing a waveform and network problem to one decimal ratio.

Another error is using phase voltage in a three-phase line-current formula that expects line-to-line voltage, or omitting the square-root-of-three factor. Delta and wye connection details matter for individual branch voltage and capacitor selection even when total balanced power formulas look similar. Keep the chosen voltage definition beside every equation.

Avoid correcting from a nameplate PF alone. Actual loading, voltage, control mode, and harmonic content can differ. Verify that any correction equipment is coordinated with generators, variable-speed drives, contactors, protection, and utility controls. Recheck the full operating range after installation rather than validating only one load point.

The simple triangle and capacitor formula assume balanced sinusoidal steady-state operation at one frequency. They do not model harmonics, interharmonics, unbalance, transients, resonance, voltage dependence, generator control, dynamic reactive support, motor starting, or protection. Detailed installations may require time-series and frequency-domain analysis.

ScholarTool provides preliminary educational relationships, not electrical design approval. Power-factor equipment can store hazardous energy and affect fault levels, switching transients, overvoltage, and power quality. Final measurement, bank or filter selection, switching, protection, conductor sizing, code compliance, and utility coordination require applicable standards, equipment data, site studies, and qualified electrical professionals.