Power Quality: The Properties Nobody Notices Until They Fail
Electricity is usually discussed as a quantity - kilowatt-hours delivered, megawatts available. It is also a waveform, and the shape of that waveform has properties that matter to the equipment receiving it. When people say the power is dirty, they are describing something specific and measurable.
What a Harmonic Is
Grid voltage is meant to be a sine wave at 50 or 60 hertz. A linear load - a resistive heater, an old filament lamp, a simple motor - draws current in the same sinusoidal shape and leaves the waveform undisturbed.
A non-linear load does not. A rectifier, a switched-mode power supply, an LED driver or a variable-speed drive draws current in short pulses, taking a gulp near the peak of each cycle and nothing in between. That pulsed current, flowing through the impedance of the supply, distorts the voltage for everyone connected nearby.
Mathematically the distorted waveform decomposes into the fundamental plus components at integer multiples - the third, fifth, seventh harmonic and so on. These carry energy that does no useful work and does cause heating. The third harmonic is particularly awkward in three-phase systems because it adds rather than cancels in the neutral conductor, so a neutral sized for balanced loads can carry more current than any phase.
The consequences are unglamorous and expensive: transformers running hot and derated, capacitors failing early, motors vibrating, and protective devices tripping for no visible reason. Because nearly every modern device is a non-linear load, total harmonic distortion has risen steadily for decades, and mitigation has become a standard part of electrical design rather than a specialist concern.
Sags, Swells and Why Industry Cares
A voltage sag is a brief drop - typically to between ten and ninety percent of nominal for anything from a half-cycle to a few seconds. The most common cause is a fault somewhere else on the network: a lightning strike or a tree contact pulls the voltage down across a wide area until protection clears it, usually within a hundred milliseconds.
For a lamp this is a flicker. For an industrial process it can be a shutdown. Contactors drop out, drives detect undervoltage and stop, and programmable controllers reset. A semiconductor fabrication plant or a continuous paper machine can lose hours of production and material to an event that lasted a tenth of a second and never registered as an outage in anyone's reliability statistics.
This is why surveys of industrial customers consistently find that sags cost more in aggregate than interruptions do. They are far more frequent, they are not compensated, and they often cannot be traced to a specific fault on the customer's own supply.
Mitigation is possible at several levels. Ride-through capability designed into the drives themselves is cheapest. Dynamic voltage restorers inject the missing voltage for a short period. Uninterruptible supplies cover the critical subset of equipment. Which is appropriate depends entirely on what a second of lost production costs, which is why the analysis is usually done by the customer rather than the utility.
Flicker, Unbalance and Reactive Power
Flicker is a slow variation in voltage amplitude, and it matters because human vision is unusually sensitive to it. A fluctuation of well under one percent at around eight to ten hertz is visible in a lamp and, for many people, uncomfortable. The classical cause is a large fluctuating load nearby - an arc furnace, a welding plant, a sawmill - and flicker limits exist principally to protect neighbours of such installations.
Unbalance occurs when the three phases carry unequal voltages, usually because single-phase loads have been distributed unevenly. Three-phase motors respond badly: the negative-sequence component that unbalance introduces produces heating and torque pulsation, and a few percent of unbalance can cost a significant fraction of a motor's rated life.
Reactive power is the most consequential and least intuitive. Inductive equipment - motors, transformers, fluorescent ballasts - draws current that is out of phase with voltage. This current does no work but still flows through every cable and transformer on the way, occupying capacity and causing losses. Networks either compensate it with capacitor banks or charge large customers for it, which is why power factor appears on industrial bills.
Voltage control across a network depends directly on reactive power flow, and this is one of the services that used to come free from large synchronous generators. As those retire, reactive support is increasingly procured explicitly - from inverters, from synchronous condensers, or from dedicated compensation equipment.
The Devices That Cause It Can Also Fix It
The useful irony of power quality is that power electronics is both the main source of distortion and the best tool for correcting it. An inverter that can synthesise a waveform can synthesise a corrective one.
An active harmonic filter measures the distortion present and injects the inverse, cancelling it near the source. A static compensator does the same for reactive power and voltage, responding within a cycle. Both are standard equipment in industrial installations and increasingly at substation scale.
More significantly, the inverters already being installed for solar, wind and storage can provide these services as a secondary function. A solar inverter sits idle every night; the same hardware can supply reactive power and voltage support through those hours at almost no additional cost. Several grid codes now require exactly this, which turns a large installed fleet into distributed power quality equipment.
What makes this work is measurement. Standards such as EN 50160 in Europe define the limits a supply must stay within for voltage magnitude, harmonics, flicker and unbalance, and compliance is established with instruments rather than argument. It is a small example of a pattern that runs through grid operation generally: the properties that matter are the ones that can be measured at a connection point, and anything connecting to a grid is judged by those numbers.
Frequently asked questions
What is power quality?
The degree to which the delivered voltage matches an ideal sine wave of correct magnitude and frequency. It covers harmonic distortion, voltage sags and swells, flicker, phase unbalance and transients - properties that are invisible to simple resistive loads and consequential for electronics and motors.
What causes harmonics?
Non-linear loads that draw current in pulses rather than smoothly: rectifiers, switched-mode power supplies, LED drivers and variable-speed drives. The pulsed current distorts the voltage for everyone nearby, producing components at multiples of the fundamental frequency that cause heating without doing useful work.
Why do voltage sags matter more than outages to industry?
Because they are far more frequent and equally disruptive. A sag lasting a tenth of a second - caused by a fault elsewhere on the network - can drop contactors and stop drives, costing hours of production, while never appearing in reliability statistics or attracting compensation.
What is reactive power?
Current drawn out of phase with voltage by inductive equipment such as motors and transformers. It does no useful work but flows through every cable on the way, occupying capacity and causing losses. Networks compensate it with capacitor banks or charge large customers for it, which is why power factor appears on industrial bills.
Can renewables improve power quality?
Yes. The inverters installed for solar, wind and storage can supply reactive power, voltage support and active harmonic filtering as a secondary function, often during hours when they are not generating. Several grid codes now require this, turning a large installed fleet into distributed power quality equipment.
