An active reactive power compensator (SVG) is most useful where reactive power demand changes rapidly, alternates between inductive and capacitive, or differs significantly between phases. With stable loads, a less expensive and simpler capacitor bank may be sufficient. The decision should be based on measurements and billing conditions, not just the latest reactive energy charge.
When is an active reactive power compensator the better option?
An SVG (Static Var Generator) is a power-electronic system that continuously adjusts the reactive current drawn from or supplied to the point of compensation. Unlike a conventional capacitor bank, it does not need to switch fixed-capacity stages in and out. It can respond to changing demand and, within its rated operating range, compensate both inductive and capacitive reactive power.
An SVG is particularly worth considering when:
- the load changes abruptly during the production cycle, for example when spot welders, overhead cranes or short-cycle equipment operate;
- the installation alternately draws inductive and capacitive reactive energy at different times or in different operating states;
- phase-to-phase differences are significant and the selected model can control each phase current independently;
- the existing stepped system responds too slowly or causes periods of undercompensation and overcompensation;
- a planned upgrade will change the load mix, requiring a wide range of stepless control.
However, none of these conditions automatically justifies purchasing an active device. With predictable, predominantly inductive loads, a properly sized capacitor bank may still be the right solution. It is useful to distinguish the general principles of reactive power compensation from the assessment of a particular SVG installation.
How does an SVG work, and what should not be assumed?
An active compensator measures currents and, depending on its architecture, voltages in the installation. A controlled power-electronic converter injects current to reduce the reactive component seen at the selected measurement point. The manufacturer’s documentation should be checked for the compensation range, control method, grid parameters and permissible current loading, among other things.
Products on the market differ in functionality. For example, TDK describes an SVG for stepless compensation and load balancing, while identifying a harmonic filter as a potentially separate component of the system. CIRCUTOR documentation, meanwhile, describes versions for three-wire and four-wire networks. This matters in installations with many single-phase loads. Details depend on the model: TDK – PQvar, CIRCUTOR – SVGm.
Dynamic control does not mean unlimited capacity
Marketing materials may quote response times in milliseconds. One specification should not be applied to every compensator. Compare exactly what the manufacturer measures: the onset of the response, the time needed to reach the set compensation level, or settling time after a load step. Available reactive power or current at the actual voltage and temperature, and while other functions operate simultaneously, also matters.
Even a very fast system will not solve the problem if it is undersized, measures the wrong point in the installation or has incorrectly configured current transformers.
What data should be collected before choosing a technology?
1. Invoices and billing terms show the scale of the problem
Start with electricity distribution invoices, ideally covering different periods of plant operation. Check inductive and capacitive reactive energy charges separately, along with the tariff group, network operator and contractual terms. Invoices help quantify costs but cannot show how quickly the load changes or the extent of phase unbalance.
In Poland, billing is governed by the Regulation of the Polish Minister of Energy of 3 September 2026 (Journal of Laws 2026, item 1182), in force since 5 September 2026, together with the applicable network operator tariff. For inductive reactive energy, the reference is the tgφ₀ coefficient specified in the connection conditions or contract; the regulation generally provides for 0.4, but the individual customer’s terms must be checked. Capacitive reactive energy is billed under separate rules. Sources: Polish Journal of Laws – 2026 regulation and Polish Energy Regulatory Office (URE) – published tariffs.
2. Active and reactive power profiles reveal how the plant operates
For an initial assessment, collect active power P and reactive power Q profiles, together with information about changes in operating modes. A technical assessment should distinguish inductive and capacitive values, examine them over successive periods and, where relevant, assess phases L1, L2 and L3 separately. Measurements at intervals shorter than 15 minutes may be needed, especially for rapidly varying loads.
A 15-minute profile is useful for cost and trend analysis, but it cannot replace recordings of short-term changes in currents and power. A quarter-hour average can hide alternating operating states that require compensation in both directions. Production schedules, photovoltaic system data and a diagram of the existing compensation system should also be collected.
If changes are caused, for example, by power converters or variable-speed drives, consider the actual operation of electric motors and drives, not just their nameplate ratings.
3. Harmonics and phase unbalance need separate assessment
In installations with inverters, rectifiers and electronic power supplies, also measure current and voltage distortion, including the harmonic spectrum and THDi and THDu indices. These are not interchangeable: one refers to current, the other to voltage. Check neutral conductor loading, phase unbalance and transformer operating conditions.
IEC 61000-4-30:2025 can be used as a reference when selecting a power quality measurement method, and IEC 61000-4-7 for harmonic analysis. These standards describe measurement methods; they do not, by themselves, require every plant to carry out an identical set of tests.
SVG, capacitor bank or hybrid system – a comparison
| Criterion | Capacitor bank | Active SVG compensator | Hybrid system |
|---|---|---|---|
| Type of compensation | Primarily inductive reactive power drawn by loads, compensated using capacitive stages | Inductive and capacitive reactive power within the device’s operating range | The passive section covers steady demand, while the active section covers variable demand |
| Load variations | Effective with stable or slowly varying loads; thyristor-switched versions can respond faster than contactor-switched banks | Stepless control, also suitable for rapid changes | Sized to the steady and dynamic components of the load profile |
| Control accuracy | Limited by the size of the switching stages | Stepless within the available range | Stepless correction of the remainder after the stages are switched in |
| Phase unbalance | Standard three-phase stages do not provide independent control of each phase | Possible with a suitably selected version | Depends on the functions of the active section |
| Harmonics | Requires an assessment of resonance risk; detuned capacitor banks are often used | An SVG alone is not automatically an active harmonic filter | Filtering functions must be specified separately |
| Costs and maintenance | Usually a simpler design; inspection of stages, capacitors and contactors | Greater complexity; internal losses, cooling and power electronics | More components, but their ratings can potentially be optimised |
First assess the steady and variable portions of reactive power demand Q. Only then can the passive, active and hybrid options be compared responsibly. Adding detuning reactors to a capacitor bank does not turn it into an active harmonic filter. The purpose of these reactors – reducing the risk of harmonic amplification and capacitor overload – is described by Schneider Electric.
SVG versus active harmonic filter – an important distinction
Reactive power compensation and harmonic filtering serve two different purposes. An SVG is primarily intended to supply the appropriate fundamental-frequency reactive current. An active harmonic filter (AHF) primarily reduces selected current harmonics. Some devices combine these functions and can also reduce phase unbalance, but the capabilities must be confirmed in the specific model’s documentation.
The term “active” or “ASVG” alone is not sufficient evidence that a device will compensate harmonics of specified orders. If several functions share the same current capacity, simultaneous operation may reduce the headroom available to each. Separate SVG and AHF modules are described, among others, by TDK; ABB also distinguishes dynamic compensation from active filtering technologies.
If the issue is reactive energy charges, first determine the required range of Q compensation. If conductors overheat, disturbances occur or THDi is high, diagnostics must also address harmonics. A single device can perform both functions only if its specifications and tests confirm this.
How should a compensator be sized without oversizing it?
Suppose that at the plant’s measurement point, one operating state produces Q = +60 kvar (using the convention that a positive value means inductive reactive power consumption), while another produces Q = −20 kvar (capacitive). If the approximate aim is to bring Q close to zero, the system would need to provide around −60 kvar in the first state and +20 kvar in the second. A conventional capacitor bank sized only for the first state could worsen overcompensation in the second.
This is an illustrative calculation, not a completed equipment specification. In a balanced 400 V three-phase system, 60 kvar corresponds approximately to the following reactive current:
I ≈ Q / (√3 × U) = 60 000 var / (1,732 × 400 V) ≈ 86,6 A.
A real design must also account for maximum and recurring instantaneous values, current distribution across phases, operating voltage, temperature, overload capability and the priorities assigned to additional functions. The control target need not be Q = 0: it should reflect the actual billing terms and technical requirements.
Specifications to require in a quotation
- compensation range in both directions and its dependence on voltage;
- rated and maximum output current, overload capability and temperature limits;
- response time with an unambiguous measurement method;
- supported network configuration (3P3W or 3P4W), independent phase control and any neutral-current compensation;
- actual harmonic-related functions, their priorities and capacity during simultaneous operation;
- internal losses, cooling, noise, installation requirements and servicing arrangements;
- communication with the supervisory system and ability to operate alongside the existing capacitor bank.
If a quotation states only “device rating: 60 kvar” and promises “compensation to cosφ = 1”, important information is still missing for a sound decision.
Design, settings and commissioning acceptance – what should be checked?
Measurement point and current transformer configuration
Current transformer placement determines which part of the installation the controller ‘sees’. Before commissioning, verify the transformer ratio, polarity, phase assignment and the equipment’s behaviour when active power flows in either direction. This is particularly important in installations with PV and multiple power sources.
It is also necessary to establish whether the control objective is to reduce Q exchange at the electricity offtake point or to serve a particular switchboard or load. These are not always the same tasks. If a capacitor bank is already installed, the two controllers must not cause unstable switching or compensation oscillations.
Acceptance testing requires before-and-after measurements, not just start-up
The acceptance report should specify the measurement point, load conditions and parameters to be checked: P, Q, inductive/capacitive operating mode, individual phase values, response to load changes and, where relevant, harmonics. The comparison should cover comparable plant operating states, including minimum and maximum load.
The technical outcome should then be checked against subsequent electricity bills. A reduction in Q at a particular moment does not yet prove that every reactive energy charge will disappear. The operator’s metering and billing rules, the plant’s load profile and other bill components all matter.
Does an active compensator always reduce energy costs?
No. Compensation can reduce reactive energy charges and unnecessary current flows in parts of an installation, but it does not automatically lead to a proportional reduction in active energy consumption in kWh. The compensator also consumes electricity to operate and requires maintenance and suitable cooling conditions.
An economic assessment should compare the actual charges that can be reduced with the costs of equipment, measurements, engineering, installation, commissioning and operation. Alternatives also need to be considered: an automatically controlled capacitor bank, a detuned capacitor bank, a reactor bank for capacitive reactive power compensation, or a hybrid solution. Detuning reactors connected in series with capacitors serve a different purpose from separate reactors used to compensate capacitive reactive power. Compare not only purchase costs but also system behaviour over a typical production cycle.
A reliable payback period cannot be calculated from a single invoice and a catalogue price list. If a charge occurs only in a few months or results from one particular operating mode, adjusting the existing system’s settings or carrying out further diagnostics may be the better first step.
How to make the decision in practice
First establish what the charge was for and whether the issue recurs over time. Then investigate the distribution and dynamics of Q, the load characteristics and harmonics. Only on that basis should you specify the required compensation range and compare quotations under the same operating conditions.
For stable loads, a capacitor bank may suffice. If demand switches rapidly between inductive and capacitive, an SVG becomes a strong candidate. If much of the demand is constant but part of it changes quickly, assess a hybrid solution. The final choice should follow from measurements and comparable acceptance criteria, not the name of the technology.
What information should you send for an initial assessment?
Describe how the plant operates, its existing compensation system and the issue shown on the invoice. You can send available P/Q profiles, harmonic measurement results, a single-line diagram or information on major loads – there is no need to assemble every document before the first contact. This will help determine whether the next step should be diagnostics, adjustments to the existing system or a comparison of technical options. Information can be submitted using the Energy Trend contact form.
Sources
- Regulation of the Polish Minister of Energy of 3 September 2026, Journal of Laws 2026, item 1182 – current legal basis for billing rules in Poland.
- Polish Energy Regulatory Office (URE) – electricity tariffs published in 2026 – source of the applicable network operator tariffs.
- TDK Electronics – PQvar static var generator – operating principles and hybrid solutions.
- CIRCUTOR – SVGm devices – design options and technical specifications.
- Schneider Electric – power factor correction and detuning reactors – applications of capacitor banks and reactors.
- ABB – power quality technologies – distinction between dynamic compensation and harmonic filtering.
- IEC 61000-4-30:2025 and IEC 61000-4-7:2002+A1:2008 – measurement methods and harmonic analysis.
