Power Factor: What It Is, How to Calculate It, Limits, and Correction

Understand power factor, how to calculate P, Q, and S, ANEEL’s 0.92 reference for Group A consumers, technical impacts, harmonics, and safe correction criteria.

Check it out!

The power factor expresses the relationship between active power effectively converted into useful work and the apparent power required by the electrical system. In its most general form, it is given by pf = P/S, where P is active power, in watts or kilowatts, and S is apparent power, in volt-amperes or kilovolt-amperes. In balanced sinusoidal systems, this relationship coincides with the well-known cos φ, associated with the angle between voltage and current.

This definition appears simple, but applying it to real installations requires care. Motors, transformers, and other inductive loads require reactive power to establish magnetic fields. Drives, rectifiers, UPS systems, switched-mode power supplies, chargers, photovoltaic systems, and BESS can introduce nonsinusoidal currents. In this scenario, true power factor does not depend only on the phase displacement between fundamental voltage and current; harmonic distortion also affects apparent power and total current.

In Brazil, ANEEL Normative Resolution No. 1,000/2021 establishes 0.92 as the reference power factor for Group A consumers, with billing criteria for excess reactive energy and demand. For Group B consumers, the same resolution does not provide for excess reactive-energy billing based on power factor. The value 0.92, however, is a regulatory reference rather than a universal engineering setpoint. An installation may need to operate with margin above this threshold while also avoiding capacitive overcompensation.

Power-factor correction, therefore, should not begin with purchasing a capacitor bank. The technically consistent process is measure, characterize the load, identify the source of reactive power and distortion, assess the network, size the compensation, design protection and switching, and validate performance after implementation. Em instalações com harmônicas relevantes, um banco de capacitores mal especificado pode inclusive agravar o problema por ressonância.

Active, Reactive, and Apparent Power: What Power Factor Really Measures

Active power P is the portion that performs useful work: it drives motors, heats resistive loads, illuminates spaces, powers processes, and converts electrical energy into other forms. It is normally expressed in kW.

Reactive power Q is associated with the periodic exchange of energy between the source and the magnetic or electric fields of certain equipment. Induction motors, transformers, and reactors, for example, require magnetization. This energy does not represent useful consumption in the same sense as active power, but it directly affects the current flowing through the installation. Its usual unit is kvar.

Apparent power S represents the total demand imposed on the system. Under classical sinusoidal conditions, it combines active and reactive components and is expressed in kVA.

For a balanced sinusoidal condition:

S² = P² + Q²

and:

pf = P/S = cos φ

This relationship explains why two installations delivering the same active power can require different currents. Consider a three-phase load that requires 100 kW. If the power factor is 1,00, the apparent power will be 100 kVA. If the power factor drops to 0,80, the same 100 kW correspond to 125 kVA.

The infrastructure does not “see” only the useful 100 kW. Transformers, cables, busbars, and protective devices must carry the current corresponding to apparent power. Therefore, low power factor can consume electrical capacity that could otherwise be used for new loads.

This relationship appears directly when assessing main switchboard, panel, and feeder capacity. An expansion analysis that considers only kW may incorrectly conclude that spare capacity exists when current and kVA are already close to the thermal limits of the assets.

How to Calculate Power Factor

The general expression is straightforward:

pf = P / S

In a sinusoidal system, the following is also used:

pf = cos φ

and reactive power can be related to active power by:

Q = P × tan φ

These equations make it possible to understand the effect of compensation.

Consider an installation with 100 kW of active power and a power factor of 0,80. The apparent power is:

S = 100 / 0.80 = 125 kVA

Reactive power is approximately:

Q = 75 kvar

If the objective is to raise the power factor to 0,95, the same 100 kW then correspond to approximately:

S = 100 / 0.95 ≈ 105.3 kVA

and the resulting reactive power is close to 32.9 kvar.

The difference between the two conditions is approximately 42.1 kvar. In a preliminary analysis of a linear load, this value represents the capacitive reactive power required to raise the power factor from 0.80 to 0.95.

The calculation, however, does not complete the design. The load rarely remains at 100 kW all the time. The installation may operate at 30%, 60%, or 100% of capacity throughout the day, and a fixed bank sized for maximum load can cause overcompensation during low-load periods.

In addition, the presence of harmonics changes the interpretation. The value obtained from cos φ describes only the angular displacement of the fundamental component. True power factor must consider the relationship between active power and total apparent power under actual waveforms.

Power Factor, cos φ, and Displacement Factor Are Not Always the Same Quantity

In circuits with sinusoidal voltage and current, power factor and the cosine of the phase angle can be treated as equivalent. This equivalence is no longer sufficient when current is distorted.

The displacement factor is related to the phase displacement between the fundamental components of voltage and current. The true power factor considers active power relative to total apparent power, incorporating the effect of distortion.

An electronic load may have a displacement factor close to 1 and still draw current with narrow peaks and high harmonic content. In this case, RMS current may be higher than suggested by simple cos φ.

IEEE 1459:2025 provides definitions for power quantities under sinusoidal, nonsinusoidal, balanced, and unbalanced conditions. This distinction is particularly useful in installations with power electronics, where diagnosis based only on cos φ can hide a relevant portion of the current.

In practice, this means an energy analyzer must be configured and interpreted correctly. Where available, it is useful to record:

total power factor;
displacement factor;
active power;
reactive power;
apparent power;
voltage THD;
current THD;
harmonic spectrum;
RMS current per phase.

The article on electrical harmonics, THD, and TDD examines this distinction among distortion, current, and system performance in greater depth.

Power Factor Is Not Demand Factor

The similarity of the names causes confusion, but the quantities describe different phenomena.

The power factor relates active power to apparent power. It is linked to how the installation uses electrical capacity and to the presence of reactive power and, under nonsinusoidal conditions, distortion.

The demand factor relates the maximum observed demand to a base of installed or planned power. It represents the simultaneity of load use.

An industrial facility may have many installed motors but operate only some of them simultaneously. Demand factor may be low. At the same time, if the operating motors are lightly loaded and draw significant magnetizing current, power factor may also be low.

The opposite is also possible: an installation may operate close to its maximum demand and still have a high power factor.

The content on demand factor and simultaneity addresses the temporal composition of loads. Power factor belongs to another layer of sizing: conversion among kW, kvar, kVA, and current.

Why Power Factor Becomes Low

The best-known cause is the presence of inductive loads.

Induction motors require magnetizing current. When they operate far below rated load, useful active power falls faster than the portion required for magnetization, and power factor can deteriorate.

Transformers also require reactive power. During low-load periods, magnetizing current can represent a more significant share of total current.

Other frequent causes include:

oversized motors;
motors operating unloaded for long periods;
lightly loaded transformers;
welding machines;
furnaces and specific electromagnetic loads;
older lighting systems with ballasts;
capacitor banks out of service;
failed stages in automatic banks;
process changes without reassessment of compensation;
electronic loads with distorted current;
varying operation between shifts.

The cause must be separated from the symptom. A bill with excess reactive charges does not by itself prove that the problem should be corrected with a larger bank. There may be a failed stage, a poorly configured controller, oversized transformation, compensation installed at the wrong point, or load variation incompatible with the existing bank.

What ANEEL’s 0.92 Threshold Means

ANEEL Normative Resolution No. 1,000/2021 establishes a reference power factor fR of 0.92 for Group A consumers.

When measurement conditions fall within the billing rules and power factor is below the reference, charges for excess reactive electrical energy and excess reactive power demand may be calculated.

The regulation uses specific measurements and formulas. Therefore, diagnosis of a charge should be based on interval meter data, meter records, bills, and the intervals actually considered by the distribution utility.

For Group B, a REN 1.000 determina que não haja cobrança de consumo de energia elétrica reativa excedente por fator de potência.

From a design standpoint, the value 0.92 should not be interpreted as an exact target. An installation oscillating between 0.90 and 0.94 may still be exposed to excess reactive charges even if the simple average appears acceptable. In practice, sufficient operating margin is normally maintained to absorb actual load variations.

At the same time, aggressively pursuing 1.00 can create another undesirable condition: capacitive power factor during low-load periods.

Why Low Power Factor Increases Current

In a three-phase system, for a sinusoidal approximation:

P = √3 × V × I × pf

With voltage and active power held constant, current is inversely proportional to power factor.

If the installation must deliver the same amount of kW but power factor decreases, current increases.

This additional current flows through transformers, busbars, cables, and protective devices. The effect is not merely “accounting”; it consumes real thermal capacity.

In addition, resistive losses are approximately proportional to I²R. If current increases, Joule losses increase with the square of current.

This does not mean every power-factor correction will produce a large reduction in active-energy bills. The benefit depends on where compensation is installed and which sections stop carrying the reactive component.

If a central bank is installed at the main switchboard, upstream current may decrease, but feeders between the main switchboard and distant motors continue to carry part of the reactive current up to the bank connection point. Compensation closer to the load can relieve more internal sections of the network but increases the number of distributed devices.

The design should determine where current reduction produces the greatest technical benefit.

What Are the Technical Impacts of Low Power Factor

Beyond possible excess reactive charges, the effects may appear as capacity limitations and performance loss.

The main impacts include:

greater utilization of transformer kVA capacity;
higher feeder current;
additional Joule losses;
greater voltage drop;
less margin for new loads;
heating of conductors and connections;
need for equipment with higher current ratings;
greater loading of busbars and main switchboards;
reduced overall efficiency of internal distribution;
excess reactive charges where applicable.

In an industrial expansion, this can be decisive. An organization may believe it needs to replace a transformer simply because current is high, when part of the capacity is being occupied by reactive power that could be technically compensated.

The opposite also occurs: an organization may try to “free capacity” using capacitors alone when the transformer is already limited by heating, harmonics, imbalance, or other factors. The diagnosis must consider the complete system.

Inductive Power Factor and Capacitive Power Factor

Inductive loads tend to demand inductive reactive power. Capacitors provide reactive power in the opposite direction and can compensate for this component.

The problem arises when compensation exceeds the load requirement.

A bank sized for maximum production may remain connected after many motors have been switched off. At that point, capacitive reactive power can exceed inductive reactive power and drive the installation into a capacitive condition.

This is common in:

fixed banks applied to variable loads;
controllers with inappropriate setpoints;
stuck stages;
process changes without bank review;
night periods or weekends;
installations that added photovoltaic generation or BESS;
transformers operating unloaded while banks remain connected.

Overcompensation is not “beneficial excess correction.” It can alter the voltage profile, interaction with the grid, and equipment behavior. The strategy should keep power factor within a coherent operating range rather than pursuing the highest possible number.

Harmonics Change Power-Factor Analysis

In installations with variable-frequency drives, UPS systems, BESS, or other nonlinear loads, the compensation proposal should be reviewed together with harmonics, resonance, protection, and operating regime. A bank suitable for sinusoidal load may be unsuitable when the current spectrum changes.

Design Review in Engineering Projects

Nonlinear loads draw current with a waveform different from voltage.

Variable-speed drives, UPS systems, rectifiers, solar inverters, EV chargers, BESS, switched-mode power supplies, and many electronic devices can generate harmonic components.

Under these conditions, power factor can be degraded even when the phase displacement of the fundamental component is small.

The conceptual relationship now involves two dimensions:

fundamental displacement;
current distortion.

Therefore, equipment may report cos φ close to 1.00 and still contribute to high RMS current.

When the facility contains relevant nonlinear loads, the correction design should analyze the harmonic spectrum before applying capacitors.

Capacitors have reactance that decreases with frequency. Therefore, they can carry significant harmonic currents. In addition, the combination of bank capacitance and equivalent network inductance has a resonant frequency.

If a significant harmonic order is close to this frequency, amplification may occur.

Resonance between Capacitors and the Network

Resonance is one of the main reasons capacitor banks should not be specified solely by kvar rating.

The network has inductance associated with transformers, cables, and other elements. The bank adds capacitance. Together, they form a circuit with a natural frequency.

When harmonic currents excite this condition, the following may occur:

overcurrent in capacitors;
increased voltage distortion;
heating;
premature failure of fuses and capacitors;
undesired protection operation;
stress on transformers;
problems in sensitive equipment.

Risk depends on system short-circuit power, bank rating, transformer impedance, and the harmonic spectrum of the loads.

In installations with significant harmonic content, a solution may require detuned banks, with series reactors, or specific active or passive filters.

The use of a reactor should not be treated as an optional accessory added “for safety.” Detuning frequency, design current, and component withstand capability need to be defined from the actual characteristics of the network.

How to Identify Whether the Problem Is Really Power Factor

If the motivation is excess reactive charges, capacitor failure, heating, or lack of capacity, the first step is to separate the symptom from the cause. Time-based measurement of power, power factor, currents, and harmonics avoids sizing compensation from an incomplete snapshot of the installation.

Power Quality Analysis and Diagnosis

The diagnosis should begin with the question that triggered the investigation.

If the motivation is excess reactive charges, invoices and utility metering data need to be analyzed.

If the problem is lack of capacity, the analysis should include current, kVA, load profile, and loading of transformers and feeders.

If there are capacitor failures, protection trips, or heating, the assessment needs to include harmonics, voltage, and the condition of the bank itself.

If the problem is voltage drop, the portion caused by high current should be separated from other factors such as cable impedance, motor starting, and source capacity.

A measurement campaign may record:

voltage per phase;
current per phase;
active power;
reactive power;
apparent power;
true power factor;
displacement factor;
voltage THD;
current THD;
harmonic spectrum;
demand;
imbalance;
events;
status of bank stages.

Measurement needs to cover representative periods. A ten-minute reading may be useless in a plant whose process changes by shift.

An industrial facility may have satisfactory power factor during production and significant overcompensation overnight. Another may present problems only during simultaneous starts or certain production cycles.

The Power Quality Analysis and Diagnosis service is appropriate when behavior needs to be characterized through measurement and correlation among power factor, harmonics, voltages, currents, and events.

Where to Measure Power Factor

The measurement point needs to match the objective of the analysis.

To compare with billing, the point of interest is the utility metering point.

To locate the source of reactive power, measurements need to move inside the facility.

A practical strategy is to work in levels:

main service entrance of the consumer unit — shows overall behavior;
main busbars — separates transformers, areas, or major load centers;
feeders — identifies process sectors;
critical loads — characterizes specific motors, machines, and equipment.

This method helps determine whether compensation should be centralized or distributed.

If a single process generates much of the reactive demand, compensating close to it may relieve internal cables and transformers.

If many small and variable loads make up the problem, a central automatic bank may be more rational.

Strategies can also be combined.

How to Analyze the Electricity Bill When There Is Excess Reactive Power

The bill is an important starting point, but it does not replace measurement.

First, determine whether the unit belongs to Group A and whether there are charges associated with excess reactive energy or reactive power demand.

Then verify:

billing period;
demand values;
reactive-energy records;
time periods or tariff periods where applicable;
recurrence of charges;
recent changes in the facility;
history before and after process changes.

If the charge began suddenly, a capacitor-bank failure may be more likely than a structural change in the entire load.

If the problem is seasonal, compensation may need greater granularity.

If on-site generation exists, the profile at the metering point may have changed even without a significant change in internal consumption.

The analysis should transform the bill into a technical hypothesis and confirm that hypothesis through measurement.

Power-Factor Correction Strategies

There is no single architecture.

Correction can be individual, group-based, or centralized.

Individual Compensation

Compensation is installed close to the specific load.

It is suitable when the equipment has predictable operation and represents a relevant share of reactive demand.

The main advantage is reducing reactive current along that load’s feeder.

The main disadvantage is multiplying maintenance points and requiring appropriate logic to prevent the capacitor from remaining connected when the load is switched off.

Group Compensation

A set of loads with similar behavior is compensated at a panel or control center.

It can provide a good balance between electrical efficiency and operational simplicity.

The design needs to confirm that the loads actually operate with sufficiently correlated profiles.

Centralized Compensation

A bank is installed on the main busbar.

This solution centralizes maintenance and control and is common in industrial and commercial installations.

The automatic bank connects and disconnects stages according to reactive-power demand.

The limitation is that downstream feeders may still carry reactive current between the loads and the busbar where the bank is installed.

Hybrid Solution

Larger facilities often combine local compensation of large motors or transformers with a central automatic bank to adjust the overall balance.

This approach can reduce internal currents without sacrificing operational flexibility.

Automatic Banks and Number of Stages

The number of stages defines compensation resolution.

A 300 kvar bank divided into three 100 kvar stages provides relatively coarse control. For a load varying slowly between 80 and 250 kvar, this can cause alternation between undercompensation and overcompensation.

A bank with smaller stages offers greater resolution but increases components, control complexity, and maintenance points.

An asymmetric sequence can also be used, with stages of different ratings.

Control logic should consider:

power-factor setpoint;
time between switching operations;
capacitor discharge time;
stage rotation sequence;
blocking failed stages;
avoiding excessive switching;
behavior under low load;
capacitive conditions;
presence of on-site generation.

Poorly configured controllers can damage contactors and capacitors through excessive switching even when total bank rating is correct.

How to Size a Capacitor Bank

When compensation requires a new bank, panel modifications, feeders, control, protection, or integration with existing distribution, the kvar calculation needs to be transformed into executable documentation coordinated with the installation.

Low-Voltage Electrical Design

For a predominantly linear load, required capacitive reactive power can be estimated by:

Qc = P × (tan φ1 − tan φ2)

where:

Qc is the bank rating;
P is active power;
φ1 corresponds to the initial power factor;
φ2 corresponds to the desired power factor.

In the 100 kW example, with initial power factor 0.80 and target 0.95, theoretical compensation is approximately 42.1 kvar.

The value is only the starting point of the design.

The Load Varies

If active power falls to 30 kW and the same fixed bank remains connected, the installation may become capacitive.

Sizing should therefore use the load curve rather than nameplate power alone.

Voltage Influences Capacitor Reactive Power

The reactive power supplied by a capacitor depends on voltage.

Voltage variations change the bank’s effective contribution and need to be considered when selecting components.

Harmonics Alter Bank Current

RMS current in capacitors may exceed the value predicted from the fundamental component when harmonics are present.

The bank needs to be specified for actual conditions.

Switching Generates Transients

Energizing capacitors can produce high transient currents.

Contactors, fuses, circuit breakers, and other devices should be suitable for capacitor switching.

The Connection Point Defines the Benefit

A central bank and a bank installed near the load may have the same kvar rating but produce different effects on internal distribution.

When implementation requires changes to panels, control, cables, protection, and integration with the installation, the Low-Voltage Electrical Design should transform the calculation into executable documentation.

Detuned Capacitor Banks and Reactors

When the harmonic spectrum makes a conventional capacitor bank unsuitable, a reactor can be connected in series with the capacitor.

The objective is to shift the resonant frequency of the assembly to a controlled region, reducing the possibility of amplification at relevant harmonic orders.

The term “detuned bank” does not mean harmonics are absent.

The reactor changes the impedance seen by frequency components and must also withstand operating current.

The specification should consider:

detuning frequency;
effective stage rating;
voltage across the capacitors;
RMS current;
losses;
heating;
tolerances;
expected harmonic level;
coordination with protection.

The design should also assess whether the source of the problem requires more specific active or passive filtering.

Active Filters and Dynamic Compensation

Active power filters use electronic converters to inject compensation currents.

Depending on the application, they may act on:

harmonics;
reactive power;
imbalance;
specific current components.

They are attractive when the load changes rapidly or distortion is significant.

Sizing is not based only on kvar; it must consider the required compensation current and harmonic content.

In facilities with drives, UPS systems, or variable electronic loads, this approach may be more appropriate than continuously increasing capacitor-bank capacity.

However, active filters also have current limits, losses, installation requirements, and maintenance needs. The decision should be technically compared with other solutions.

Why Power Factor 1.00 Should Not Be Pursued at Any Cost

In theory, pf = 1 represents the maximum ratio between active and apparent power.

In actual operation, maintaining exactly 1.00 at all times is difficult and not always desirable.

Loads change. Banks have discrete stages. Controllers have delay. Active power can fall quickly.

If the setpoint is very close to 1.00 and compensation is aggressive, the installation may switch into a capacitive condition.

In addition, in distorted systems, correcting only the fundamental reactive component may not eliminate harmonic current.

The objective should be a stable range that:

meets the regulatory reference;
maintains margin against variations;
avoids undesired capacitive conditions;
minimizes current where this provides benefit;
does not introduce resonance;
is compatible with load dynamics.

The best setpoint results from measurement and actual facility behavior.

Power Factor in Motors

Induction motors deserve specific attention.

At rated load, many motors have a reasonable power factor. At low load, active power delivered to the shaft decreases, but magnetization remains necessary.

For this reason, oversized motors can contribute to low power factor.

Before installing compensation, assess:

actual loading;
operating regime;
operating hours;
possibility of resizing the motor;
variable-frequency-drive control;
starts and stops;
benefit of local compensation.

Correction should not mask an operational-efficiency problem.

A 100 kW motor operating continuously at 25 kW may justify a broader investigation than simply adding capacitors.

Power Factor in Transformers

Transformers have magnetizing current and no-load losses.

When very lightly loaded, their reactive contribution may become significant relative to the active power transferred.

This occurs in facilities with many energized transformers and low utilization outside production hours.

An efficiency strategy may involve evaluating:

loading;
number of energized transformers;
parallel operation;
no-load losses;
load transfer;
required reliability.

De-energizing unused transformers may provide greater benefit than correcting power factor alone, provided the strategy is compatible with continuity and operation.

Power Factor with Variable-Frequency Drives

Modern drives may have high displacement factor at the input, but current may contain harmonics depending on topology.

Six-pulse rectifiers, for example, generate a characteristic spectrum and may require distortion analysis.

Installing capacitors on the same busbar should consider this interaction.

In addition, conventional power-factor-correction capacitors should not be connected to the output of a drive without specific assessment by the manufacturer and for the application. The PWM waveform and control dynamics make this interface different from a conventional sinusoidal supply.

The focus should be power factor and power quality at the appropriate point in the system.

Power Factor with Photovoltaic Generation

Photovoltaic generation can significantly alter the power factor observed at the point of connection.

Consider an industrial facility consuming 500 kW and 200 kvar inductive. If a photovoltaic system begins supplying a large share of the kW during the day, active power imported from the grid decreases. If the installation’s reactive demand does not fall by the same proportion, the P/S ratio seen by the meter may change.

A capacitor bank originally sized for operation without generation may become excessive during certain periods.

In addition, photovoltaic inverters may have reactive-power control capabilities according to their limits and connection requirements.

The strategy should consider:

internal load;
instantaneous generation;
inverter capability;
control mode;
utility requirements;
existing bank;
harmonics;
export or low-import scenarios.

Power Factor with BESS

Battery energy storage systems also alter power flow.

During charging, the BESS acts as a load. During discharge, it reduces imports or may export active power.

The PCS converter may be capable of supplying or absorbing reactive power within its operating envelope.

This creates control opportunities but also increases complexity.

The existing capacitor bank should not be assessed in isolation. Scenarios need to be simulated or measured:

BESS charging at maximum power;
BESS discharging;
BESS inactive;
high photovoltaic generation;
low load;
high load;
contingency operation.

The content on power quality with inverters, BESS, and distributed generation addresses these interfaces.

Power-Factor Correction and Transformer Capacity

Compensation can reduce apparent power flowing upstream of the point where it is installed.

This may free capacity margin in transformers.

But this conclusion needs to be verified by measurement.

If the transformer is limited by:

ambient temperature;
ventilation;
harmonics;
imbalance;
cyclic overload;
aging;
protection constraints;
downstream busbar capacity;

power-factor correction may not solve the main bottleneck.

For expansions, it is useful to compare scenarios before and after compensation.

The analysis should separate theoretical electrical capacity, actual thermal limitation and recommended operating margin.

Power-Factor Correction and Voltage Drop

Voltage drop depends on current and circuit impedance.

Reducing reactive current can decrease voltage drop in certain feeders.

The effect is more relevant in long or heavily loaded lines.

However, capacitors also affect the voltage profile and can raise local voltage, especially under light load.

Therefore, sizing should assess both peak-load and minimum-load conditions.

It is not enough to verify that voltage improves during production hours. It is necessary to confirm that it remains within acceptable limits under all operating regimes.

Protection and Switching of Capacitor Banks

Capacitors store energy and require specific protection.

A bank should consider:

overcurrent protection;
fuses or devices suitable for the stages;
discharge after disconnection;
contactors suitable for capacitive switching;
overvoltage protection where applicable;
ventilation;
temperature;
accessibility;
identification;
coordination with the panel.

Automatic banks also require switching logic and cycle control.

Energizing a stage can generate high inrush current, especially when other capacitors are already energized on the busbar.

The switching device needs to be selected for this service.

Discharge resistors should ensure safe reduction of residual voltage within the design time.

Capacitor-Bank Maintenance

A capacitor bank is not a “install and forget” device.

Common failures include:

open fuses;
inoperative stages;
worn contactors;
swollen capacitors;
heating at connections;
blocked ventilation;
incorrectly installed sensors or CTs;
controller with inappropriate setpoint;
excessive switching;
overheated reactors;
degradation caused by harmonics.

An inspection should combine visual assessment, electrical measurements, and functional verification.

Electrical thermography can identify abnormal heating in connections, contactors, fuses, and reactors, but it does not replace measurement of kvar per stage or power-quality analysis.

How to Validate Whether the Correction Worked

Correction is complete only when performance is demonstrated under relevant load regimes. Stage current, switching logic, power factor, harmonics, and heating need to be recorded against defined acceptance criteria.

Commissioning and Technical Acceptance of Electrical Installations

The installation should not be considered corrected merely because the controller displays 0.98.

Validation needs to demonstrate performance under different operating regimes.

An acceptance sequence may include:

verification of installation;
verification of protection devices;
measurement of each stage current;
confirmation of effective reactive power;
testing of switching logic;
measurement at minimum load;
measurement at medium load;
measurement at maximum load;
verification of inductive and capacitive conditions;
recording THD before and after;
observation of relevant transients;
confirmation of absence of abnormal heating;
review of controller parameters;
comparison against design criteria.

When the bank includes reactors or filters, acceptance should also verify the harmonic condition defined in the design.

The Commissioning and Technical Acceptance of Electrical Installations organizes this validation into traceable evidence instead of relying only on the controller indication.

How to Determine Whether the Existing Bank Is Undersized, Oversized, or Faulty

Behavior over time provides clues.

A bank that is undersized may keep all stages connected and still leave power factor below target during high-load periods.

A bank that is oversized may operate with only a few stages and still drive the installation into a capacitive condition under low load.

A bank that is faulty may appear to have sufficient installed capacity while some stages fail to provide the expected kvar.

Measurements should compare:

nominal kvar per stage;
expected current;
measured current;
fuse status;
controller command;
voltage;
temperature;
operating hours;
failure history.

Simply inspecting the panel label does not indicate how much compensation is actually available.

When to Review Power Factor

The analysis should be revisited whenever there is a relevant change in the installation.

Typical triggers include:

production expansion;
installation of new motors;
removal of machines;
shift changes;
transformer replacement;
new drives;
new UPS systems;
photovoltaic generation;
BESS;
electric-vehicle chargers;
new furnaces or welding equipment;
main switchboard refurbishment;
change in energy contract;
recurring capacitor failures;
increase in harmonics;
unexpected excess reactive charges;
topology changes.

A review is also advisable after long periods without bank maintenance.

The installation may have changed gradually while compensation remained unchanged.

How to Procure a Power-Factor Diagnosis

A diagnostic scope should clearly define the problem and expected evidence.

The scope may include:

invoice analysis;
installation survey;
identification of existing banks;
measurement points;
measurement-campaign period;
recorded quantities;
correlation with production regime;
true power factor;
displacement factor;
active, reactive, and apparent power;
harmonics;
transformer loading;
cause diagnosis;
prioritized recommendations.

The report should explain why the power factor is inadequate, not merely report the measured value.

It should also separate operational measures, maintenance of the existing bank, and the need for new investment.

How to Procure a Correction Design

When the solution requires intervention, the design should transform the recommendation into an executable scope.

Deliverables may include:

calculation memorandum;
load curve;
total compensation rating;
stage division;
location;
single-line diagram;
capacitor specification;
reactor specification, where applicable;
switching devices;
protection;
cables and busbars;
control logic;
ventilation;
installation requirements;
commissioning criteria;
initial controller parameters.

In installations with harmonics, the design should demonstrate how bank-network interaction was considered.

When the design is supplied together with the equipment, an independent review can be useful to verify assumptions, calculations, and interfaces.

The Design Review in Engineering Projects service can support this stage, especially in critical facilities or those with a strong presence of power electronics.

What to Require at Implementation Acceptance

Acceptance should be defined before procurement.

Useful criteria include:

Criterion	Expected evidence
Installed capacity	identification of stages and measured values
Protection	devices installed according to the design
Switching	validated sequence and timing
Power factor	performance recorded under representative operating regimes
Harmonics	before/after comparison where applicable
Temperature	absence of abnormal heating
Documentation	updated diagrams, settings, and manuals
Safety	compatible discharge, identification, and access
As-Built	final configuration recorded

Acceptance based on “it worked when switched on” is insufficient.

The system needs to demonstrate correct behavior across the expected conditions.

Relationship between Power Factor and Power Quality

Power factor is only one dimension of electrical performance.

An installation may have pf = 0.98 and still present:

high harmonic distortion;
voltage sags;
imbalance;
flicker;
transients;
overvoltages;
grounding problems.

The reverse may also occur: inadequate power factor without significant harmonic problems.

The investigation should avoid reducing “power quality” to a single number.

PRODIST Module 8 addresses several power-quality phenomena. The diagnosis should select indicators consistent with the actual problem.

When there are intermittent failures, trips, heating, or sensitive equipment, the article when to investigate a company’s power quality helps distinguish signs that justify a measurement campaign.

Relationship between Power Factor and Energy Efficiency

Improving power factor can reduce losses in parts of the internal distribution system, but it should not be presented as synonymous with overall energy efficiency.

Compensation does not directly reduce the mechanical power required by a motor or eliminate internal load losses.

It can reduce current upstream of the compensation point and thereby reduce resistive losses in that section.

The benefit depends on:

initial power factor;
current;
feeder length;
conductor resistance;
bank location;
load profile;
operating hours.

In some cases, loss savings are relevant. In others, the main benefit is avoiding excess reactive charges or freeing capacity.

The economic analysis should separate these components.

Relationship between Power Factor and Load Expansion

Before installing new machines, it is useful to assess whether part of the apparent-power capacity is occupied by reactive power.

If a 1,000 kVA transformer delivers 700 kW at a power factor of 0.70, apparent power is already close to 1,000 kVA.

If the load can operate stably at pf 0.95, the same active power would require about 737 kVA.

This does not automatically mean that the 263 kVA difference is “available” for any new load. Current, heating, harmonics, protection, busbars, and operating regime need to be verified.

But the analysis demonstrates why power factor can influence CAPEX decisions.

A well-planned expansion should simultaneously assess demand, power factor, transformation capacity, selectivity, and power quality.

Final Considerations

Power factor is a simple relationship only in its introductory form. In a real installation, it connects active power, reactive power, apparent power, current, losses, transformer capacity, voltage drop, billing, harmonics, capacitor banks, and dynamic load behavior.

ANEEL’s reference of 0.92 for Group A units is relevant, but it does not exhaust the engineering analysis. The objective should not be merely to avoid a charge; it should be to keep the installation stable, with compatible current, operating margin, no overcompensation, and control of risks introduced by the solution itself.

Capacitor banks are an important tool but not an automatic answer. In facilities with variable or nonlinear loads, it may be necessary to combine measurement, automatic banks, detuning reactors, active filters, converter-based control, or operational adjustments.

The correct sequence remains: measure, understand, size, design, implement, and validate. When this order is respected, correction ceases to be a reactive intervention driven by the bill and becomes part of technical management of the electrical infrastructure.

Technical References

[1] BRAZILIAN ELECTRICITY REGULATORY AGENCY. ANEEL Normative Resolution No. 1,000, December 7, 2021. Section VIII — Power Factor and Excess Reactive Energy. Available at: https://www2.aneel.gov.br/cedoc/pubren20211000.pdf

[2] BRAZILIAN ELECTRICITY REGULATORY AGENCY. Distribution Rules and Procedures — PRODIST. Brasília, DF: ANEEL. Available at: https://www.gov.br/aneel/pt-br/centrais-de-conteudos/procedimentos-regulatorios/prodist

[3] BRAZILIAN ELECTRICITY REGULATORY AGENCY. Quality of Electricity Supply. Brasília, DF: ANEEL. Available at: https://www.gov.br/aneel/pt-br/assuntos/distribuicao/qualidade-do-fornecimento-de-energia-eletrica

[4] IEEE. IEEE Std 1459-2025 — IEEE Standard Definitions for the Measurement of Electric Power Quantities Under Sinusoidal, Nonsinusoidal, Balanced, or Unbalanced Conditions. Available at: https://standards.ieee.org/ieee/1459/7578/

Frequently Asked Questions

What Is Power Factor?

It is the relationship between active power and apparent power of an installation or load. Under simple sinusoidal conditions, it equals the cosine of the angle between voltage and current. Under harmonic conditions, true power factor is also influenced by distortion.

What Is ANEEL’s Minimum Reference Power Factor?

ANEEL Normative Resolution No. 1,000/2021 establishes a reference power factor of 0.92 for Group A consumer units, with billing rules for excess reactive power. For Group B units, the resolution does not provide for billing excess reactive energy based on power factor.

Does a 0.92 Power Factor Mean the Installation Is Technically Optimized?

Not necessarily. 0.92 is a regulatory reference for Group A. The installation may require a greater operating margin or present current, harmonic, capacity, or overcompensation issues that need to be assessed separately.

How Do You Calculate a Capacitor Bank for Power-Factor Correction?

For predominantly linear loads, Qc = P × (tan φ1 − tan φ2) can be used. The result is only a sizing basis: load profile, voltage, harmonics, location, switching, and protection need to be considered in the final design.

Does a Capacitor Bank Solve Harmonics?

No. Capacitors correct reactive power but can interact with network inductance and create or worsen resonances. Installations with harmonics may require detuning reactors, filters, or other strategies defined by study.

What Is the Difference between Power Factor and Demand Factor?

Power factor relates active and apparent power. Demand factor relates actual demand to a base of installed or expected power and represents simultaneity of load use. They are different quantities.

When Should Power Quality Be Measured before Installing Capacitors?

It is advisable when nonlinear loads, inverters, UPS systems, drives, BESS, distributed generation, recurring capacitor failures, current or voltage distortion are present, or when the load profile varies significantly.

Is It Better to Correct Power Factor to 1.00?

There is no universal rule. Aggressively pursuing 1.00 can cause overcompensation under light load. The target should consider the regulatory reference, operating variations, stage resolution, and harmonic behavior of the installation.

Complementary Technical Materials