Short answer: Reactive power penalty is reflected in the bill when the ratio of inductive or capacitive reactive energy drawn or supplied to the grid (kVArh) to active energy (kWh) exceeds a certain limit during a billing period. For distribution-connected installations with a power of 50 kVA and above, the limit is 20% inductive and 15% capacitive (33% and 20% below 50 kVA); if the limit is exceeded, the entire measured reactive energy in that direction is billed. If the penalty recurs, determine the cause through measurement before upgrading the panel.
What is reactive energy? Inductive, capacitive, and cos φ.
Motors, transformers, and ballasts draw non-working power from the grid to create a magnetic field. Legislation defines this power as the power with a 90-degree phase difference between current and voltage, measured in kVAr; reactive energy is measured in kVArh (Electricity Market Consumer Services Regulation, Article 4). In inductive loads, the current lags behind the voltage; capacitors and long cables under light loads behave capacitively.
At the fundamental frequency, the ratio of reactive power to active power is tan φ. Accordingly, a 20% inductive ratio corresponds to an average cos φ ≈ 0.98, a 33% inductive ratio ≈ 0.95, and a 15% capacitive ratio approximately 0.989 capacitive (cos φ = 1/√(1 + ratio²)). This is a period average; the instantaneous value may deviate in either direction.
Reactive penalty limits: what does the legislation say?
Most facilities are connected to the distribution network. The Communiqué on the Regulation of Distribution Tariffs leaves it to the distribution companies to determine to whom and how the reactive energy charge will be applied (Articles 4, 5/4 and 33). The rules are set by the Energy Market Regulatory Authority (EPDK) in its text ’Procedures and Principles Regarding Tariff Applications of Distribution License Holder Legal Entities and Authorized Supply Companies’; the reactive provisions are in Article 13, which was rewritten in 2022 with Board Decision 10961, and the calculation method is in Annex I.
Reactive energy tariffs do not apply to residential, lighting, and single-phase consumers, and users with an installed capacity of up to 15 kW (including 15 kW) (Article 13/1). For others, the limit depends on the installed capacity: if the ratio of inductive reactive energy drawn to active energy exceeds 33% below 50 kVA, and 20% above 50 kVA; and if the ratio of capacitive reactive energy supplied exceeds 20% or 15% respectively, a reactive energy charge is incurred (Articles 13/2 and 13/3, Appendix I).
If the limit is not exceeded, there is no charge. If it is exceeded, not only the excess amount but the entire reactive energy measured in that direction is billed (Article 13/4, Appendix I); if both limits are exceeded, the reactive meter recording the highest value is taken as the basis (Article 13/6). The cost of the first violation in the calendar year is shown on the invoice and deducted on the same invoice; all subsequent violations are payable. For readings exceeding 35 days, the charge is limited to 35 days (Article 13/1). If reactive meters do not record any value due to user error, billing is done based on reactive energy equal to 0.90 times the active energy (Article 13/7).
For consumers directly connected to the transmission system and companies holding distribution licenses, the Electricity Network Regulation applies: the ratio of inductive reactive energy drawn monthly to active energy drawn cannot exceed 20%, and the ratio of capacitive reactive energy supplied cannot exceed 15% (Article 14/1); sanctions are regulated in the connection and system usage agreements (Article 14/4).
| User or situation | Provision | Rest |
|---|---|---|
| Residential, lighting, single-phase; installed capacity up to 15 kW (inclusive) | Reactive energy tariff does not apply. | Procedures and Principles, Article 13/1 |
| Distribution-connected, installed capacity below 50 kVA. | Charge applies if inductive > 33% or capacitive > 20% | Article 13/2, Appendix I |
| Distribution-connected, installed capacity of 50 kVA and above. | Charge applies if inductive > 20% or capacitive > 15% | Article 13/3, Appendix I |
| When the limit is exceeded | The entire reactive energy measured in the crossed direction; the highest meter reading if both limits are exceeded. | Articles 13/4 and 13/6 |
| First breach of the calendar year | The fee is deducted from the same invoice; subsequent violations are paid. | Article 13/1 |
| Consumer directly connected to the transmission, distribution company | Inductive ≤ 20%, capacitive ≤ 15%; not assessed in a month when the monthly active consumption at the busbar is below 500 MWh | Network Regulations Article 14/1 and 14/3 |
These are the official regulations we could find as of November 2026; for unit prices, please refer to the EPDK's current tariff tables and the reactive energy charge line on your bill.
Reactive power penalty calculation: find your rate from your invoice.
The meter reading report includes active and reactive indices, the meter multiplier, and current and voltage transformer ratios, if applicable (Consumer Services Regulation, Article 32/6). Multiply the index difference by the multiplier to calculate the two ratios:
- Inductive ratio (%) = drawn inductive kVArh ÷ drawn active kWh × 100
- Capacitive ratio (%) = supplied capacitive kVArh ÷ drawn active kWh × 100
| Hypothetical example (both above 50 kVA) | A: two-shift factory | B: Office building |
|---|---|---|
| Monthly active energy | 120,000 kWh | 40,000 kWh |
| Inductive reactive energy | 27,600 kVArh | 2,000 kVArh |
| Capacitive reactive energy | 4,800 kVArh | 7,200 kVArh |
| Inductive/capacitive ratio | %23 / %4 | %5 / %18 |
| Result against the 20% and 15% limits | Inductive overload | Capacitive overload |
| Reactive energy as the basis for the fee. | 27,600 kVArh | 7,200 kVArh |
In facility A, the inductive limit is 24,000 kVArh, with an overshoot of 3,600 kVArh. Assuming approximately 400 operating hours per month, this represents an average deficit of 9 kVAr; such a deficit often indicates a decommissioned stage rather than a new circuit breaker. In building B, assuming capacitive energy accumulates during approximately 500 hours at night and on weekends, a continuous capacitive base load of around 14 kVAr is obtained.
The amount is the product of the reactive energy used as the basis for the charge and the unit price of reactive energy (Appendix I). Since the limit has been exceeded, the entire measured value is taken as the basis (Article 13/4): Facility A pays based on 27,600 kVArh instead of 3,600, and building B pays based on 7,200 kVArh. If this is the first violation in the calendar year, the amount is deducted from the same invoice; payment is made from the second violation onwards.
Relating to: How is the Reactive Power Penalty Calculated?
Why are reactive power penalties issued in modern facilities?
Capacity overload: night and weekend
We are encountering capacitive overload more and more frequently in our measurements. LED drivers, UPS (uninterruptible power supply) and input filters of computing loads behave capacitively; long cables also provide capacitive power under light load. Motor load during the day covers this, while capacitive energy accumulates at night and on weekends. Typical causes include fixed capacitors left in the circuit, a first stage larger than the night load, and relays that release the stage too late. IEC 61921 also recommends that capacitors be partially or completely removed from the circuit because the voltage rises under light load (clause 6.2.1).
Inductive overload: there's a control panel but it's not holding the step.
The relay stages show that they are active, but stages with blown fuses, non-closing contactors, or decapacitated capacities will not supply power. The target cos φ, C/k (relay stage sensitivity setting), and delay times may be incorrect. The most frequently overlooked point is the current transformer: it must be on the grid side of the compensation connection to see both the total load current and the compensation current; incorrect phase or reverse polarity will mislead the relay.
kVAr calculation: How many kVAr of capacitor are needed?
The required compensation power Qc is found using the formula Qc = P × (tan φ1 − tan φ2). Hypothetical example: In a facility with an average load of 400 kW, if the power factor is to be increased from 0.80 to 0.98, then tan φ1 = 0.75, tan φ2 ≈ 0.203, and Qc = 400 × (0.75 − 0.203) ≈ 219 kVAr; in practice, a bank of around 220 kVAr is considered. The total power is selected according to the highest measured reactive power demand, and the smallest step is selected according to the lowest load.
Dividing 220 kVAr into 10 + 20 + 40 + 50 + 50 + 50 kVAr stages allows you to reach all intermediate values in 10 kVAr increments. If the night load remains below 10 kVAr, the problem is not in the number of stages, but in the size of the smallest stage.
Relating to: How many kVAr capacitors are needed for compensation?
Compensation in harmonic grids: why reactors?
Variable speed drives, UPS systems, and LED drivers draw non-sinusoidal currents; in such a network, a capacitor without a reactor forms a resonant circuit with the transformer inductance. The harmonic order of the resonance is approximately given by n ≈ √(S/Q), where S is the short-circuit power at the connection point and Q is the capacitor power (IEC 61921:2017, Appendix B.2).
Hypothetical example: A transformer with a power of 1,600 kVA and a short-circuit voltage of 6% has a short-circuit power of approximately 26.7 MVA at its low-voltage (LV) busbar, ignoring the upper grid. When the bank is operating at 300 kVAr, the resonance is approximately at the 9.4th harmonic, at 500 kVAr at 7.3, and at 800 kVAr at 5.8. As the stages change, the resonance shifts over the 5th and 7th harmonics, predominantly produced by six-pulse drives; total harmonic distortion (THD) and capacitor current increase.
IEC 61921 states that unless a suitable tuning reactor is installed in series at each stage, the bank will generally amplify harmonic amplitudes and shorten capacitor life (clauses 6.3.1.1 and 6.3.2.1). According to IEC 61642, the most common solution is to lower the tuning frequency below the least significant harmonic using a series reactor. The reactor ratio is p = XL/XC, and the tuning order is 1/√p. If the least significant harmonic is the 5th, it is recommended to choose p above 4%; if there is a strong 3rd harmonic between phases, p should be above 11% (IEC 61642:1997, clause 2.5). With a 7% reactor, the tuning frequency on a 50 Hz network will be approximately 189 Hz.
The reactor increases the capacitor voltage by approximately 1/(1 − p) times at the fundamental frequency; in a 7% to 400 V grid, the capacitor experiences approximately 430 V, and the rated voltage is selected accordingly (IEC 61921:2017, clause 6.2.1). Therefore, it is often not possible to add a reactor to an existing flat capacitor; mixing reactor and non-reactor groups can also lead to an unbalanced distribution of harmonic load (IEC 61642:1997, clause 3.3).
The final project must include compensation calculations, total load current calculations for harmonic loads, resonance frequency calculations, and determination of the reactor/capacitor ratio (Regulation on the Preparation of Electrical Installation Projects, Article 10/c-4). In a network with very low harmonic content and high resonance frequency, a flat capacitor may be sufficient (IEC 61642:1997, Article 3.2); this will be demonstrated by the measurement.
Relating to: Harmonic Filter Selection: Passive, Active, and Reactor-Based Solutions
Shunt reactor in capacitive problem: only if measurement shows otherwise.
A shunt reactor is a winding that draws reactive power from the busbar to which it is connected (Electrical Network Regulations, Article 4). It is not the first solution for capacitive penalty: first, fixed capacitors, the smallest tap, and the tap release settings of the relay are corrected. If a continuous capacitive base load remains during night measurement (such as 14 kVAr in building B), a fixed or relay-controlled shunt reactor sized according to this value makes sense; an unnecessarily large one will create inductive overload during the day.
Frequent malfunctions and checks in the compensation panel.
Capacitors operate continuously at full load while in the circuit; overvoltage, overcurrent, and overheating shorten their lifespan (IEC 61921:2017, clause 6.1). Common situations we encounter in the field include:
| Symptom | Possible cause | Control |
|---|---|---|
| Swollen capacitor, frequently blown fuse. | Overvoltage, harmonic current, resonance, high panel temperature. | THD measurement, ventilation, resonance control. |
| Blackened or stuck contactor | Incompatible contactor with capacitor, faulty pre-charge resistor, frequent switching. | Contactor type, precharge resistors, relay delays |
| Heating in the busbar, terminal or reactor. | Loose connection, overcurrent, reactor saturation | Thermography under load |
IEC 61921 requires that connections to switching and protective elements continuously carry at least 1.3 times the rated current; it recommends regular inspection of contacts as poor contact can cause arcing and stress the capacitor (clauses 6.2.2 and 6.3.1.2).
The actual power from the measured capacitance is found using the relationship in IEC 61921 Annex B.1: Q ≈ ⅔ × (Ca + Cb + Cc) × ω × U² with three measurements between lines. Hypothetical example: at 400 V, each measurement should be approximately 249 µF for a capacitor labeled 25 kVAr; if 224 µF is measured, the actual power is approximately 22.5 kVAr. The manufacturer's tolerance is taken into account when making a replacement decision.
Why a week's worth of measurements before enlarging the dashboard?
When a penalty is imposed, the first reflex is to buy a larger panel. However, the problem is often not insufficient power but incorrect power: in case of capacitive overload, a large panel worsens the situation, and in a harmonic network, an additional stage without a reactor shifts the resonance to another harmonic. A week's worth of recordings shows shifts, nights, and weekends together. The following is recorded in the main LV panel:
- active, inductive and capacitive reactive power profile,
- Step behavior and relay alarms,
- current and voltage harmonics,
- voltage level and imbalance,
- Panel and capacitor temperatures (thermography).
The result might be repair and adjustment, a change in the stage structure, or investment in reactor compensation, filters, or shunt reactors. Clearly stating that no investment is needed is also a result.
Relating to: What is Energy Analysis and How is it Performed?
Frequently Asked Questions
What is the reactive penalty limit percentage?
For distribution-connected users with an installed capacity of 50 kVA and above, the limits are 20% inductive and 15% capacitive; below 50 kVA they are 33% and 20% (Procedures and Principles on Tariff Applications, Art. 13). No reactive energy charge applies to residential, lighting and single-phase users or to users with an installed capacity of up to 15 kW. For consumers connected directly to the transmission system, the 20% and 15% limits of the Electricity Grid Regulation apply (Art. 14).
What is the formula for calculating reactive penalty?
The rate is the ratio of inductive or capacitive kWh to active kWh. If the limit is exceeded, the amount is the product of the total reactive energy measured in the direction of the excess and the unit price; if both limits are exceeded, the higher value is taken (Articles 13/4 and 13/6, Appendix I).
What should cos φ be?
A 20% inductive ratio corresponds to a period-average cos φ of about 0.98. The relay target is usually set to 0.98–0.99 inductive; pushing it to 1.00 or into the capacitive region can cause a capacitive charge at low load.
Will a penalty still be issued in case of compensation system failure?
Yes. The cost depends on exceeding the limit; there is no separate exception for compensation failure in Article 13 of the Procedures and Principles. However, the cost of the first violation in a calendar year is deducted from the same invoice, while subsequent violations are charged (Article 13/1). Consider the first violation as a warning and fix the fault immediately; if you monitor the monthly indices, you will catch the level fault before it appears on the invoice.
Sources
- Electricity Grid Regulations, Articles 4 and 14 (compensation of reactive energy) — Official Gazette 28/5/2014, number 29013
- Procedures and Principles Regarding Tariff Applications of Distribution License Holders and Authorized Supply Companies, Article 13 and Annex I (Amended by EPDK Board Decision 10961) — Official Gazette, April 30, 2022, No. 31825
- Circular Regarding the Regulation of Distribution Tariffs, Articles 4, 5 and 33 (reactive energy charge) — Official Gazette 19/11/2020, number 31309
- Electricity Market Consumer Services Regulation, Articles 4 and 32 (definitions, reading notice) — Official Gazette 30/5/2018, number 30436
- Regulation on the Preparation of Electrical Interior Installation Projects, Article 10 (content of the compensation project) — Official Gazette, 3/12/2003, number 25305
- IEC 61921:2017, Power capacitors – Low voltage power factor correction banks
- IEC 61642:1997, Industrial AC networks affected by harmonics – Filter and shunt capacitor application.
Relating to: Engine Compensation Calculation
Relating to: Compensation Systems and Reactive Power Control
Relating to: What are Harmonics? THD, TDD, and Limit Values
Relating to: Power Quality Measurements
If you would like to schedule measurements for reactive power penalties or compensation failures at your facility, please contact us. For evaluation, a week's worth of data from the main distribution panel and the bills from the last twelve months are usually a sufficient starting point.