top of page

Blog Posts

Installation and Commissioning of Automatically Controlled Compensation Banks

In recent decades, manufacturers of automatic compensation banks have encountered heightened global competition, necessitating cost-effective production methods. Presently, advancements in technology have led to the development of more compact and lighter capacitors compared to those of the past.

The installation and commissioning of automatically controlled compensation banks have been facilitated by the minimization of capacitors. This advancement has led to the creation of steps—modules equipped with capacitors, discharging resistors, fuses, contactors, and, if necessary, reactors—all housed within standardized industrial enclosures.

Power factor relays are typically installed in the doors. With the reduced active power losses in the capacitors, it is now feasible to assemble compensation banks of 400 kvar or more within a single cubicle measuring (W × H × D) 600 mm × 2000 mm × 400 mm, without the need for reactors.

Let's now discuss the installation and commissioning of low-voltage automatically controlled compensation banks:

1. Installation Requirements

Initially, it is essential to consult the national guidelines concerning the installation of electrical power plants with a rated voltage up to 1000V. European standards, specifically EN 60931-1 and 3, set forth the regulations for installing capacitor banks, which are applicable in numerous countries globally.

The customer is responsible for the correct selection and installation of electrical components, with assistance from technical specialists.

Before the installation of a new automatically controlled compensation bank, it is necessary to verify that the bank's technical specifications meet the criteria outlined in the individual orders, for example:

  1. Is the rated voltage the same as or higher than the nominal voltage of the grid?

  2. Is the rated reactive power adequate for the application and can it be extended if necessary?

  3. Does the allowable ambient temperature match the actual one?

  4. If the reactor is protected, does its rate match the one specified in the order?

  5. Does the integrated power factor controller meet the input data regarding the external current transformer, and are there free steps for expansion?

  6. Is there enough space to install the bank, ensuring it's dry and away from heat sources?

  7. Is the installation site free from vibrations, or otherwise, is the bank set on vibration dampers?

  8. Is the technical data of the compensation bank readily visible at all times?

When selecting power cables for a compensation bank, their cross-section should be based on a value higher than the nominal current. As per European standards EN 60831-1 for LV capacitors[^5] and EN 60871-1 for MV capacitors[^2], the cables must be designed to carry 1.3 times the nominal current continuously.

Additionally, a positive tolerance of up to 10% must be factored into the capacitor's capacitance. These factors (1.3 × 1.1) lead to an overcurrent of Io = 1.43 × In. Therefore, it is advisable to select the cross-section of cables and conductors at approximately 1.5 times the nominal current.

The overload capability, along with the high inrush current to the capacitors, should be considered when designing protective devices and determining cable cross-sections.

Table 1 provides a list of nominal currents, cable cross-sections, and fuses associated with capacitors or banks of varying nominal reactive powers suitable for grid nominal voltage levels of 230, 400, and 525V.

For example, a compensation bank rated for a reactive power of 100 kvar at a nominal voltage of 400 V would have a nominal current of 144 A. Multiplying this by 1.43 yields an approximate current of 200 A. Consequently, as per Table 1, a cable cross-section of 3 × 95/50 mm2 would be required.

Principal circuit diagram of an automatically controlled compensation bank
Figure 1 – Principal circuit diagram of an automatically controlled compensation bank (NSP = Low-voltage distribution panel)

2. Setting into Operation of Automatic Compensation Banks (ACBs)

Figure 1 presents the main circuit diagram of an automatic compensation bank. The unit, as supplied by the manufacturer, is fully wired and prepared for connection.

The primary task of electricians is to connect the bank to the distribution plant using the selected power cables. The supplied fuses or load circuit breakers must adhere to the specifications mentioned previously.

Additionally, a cable with a minimum cross-section of 2×2.5 mm2 should be installed from the measuring current transformer to the power factor controller in the automatic compensation bank. The required cross-section will vary based on the distance between them, as depicted in Figure 2a and 2b, which illustrate options for current transformers with nominal secondary outputs of 5 A and 1 A, respectively.

2.1 Selection of CT and determination of the CT cable

Most current electrical distribution plants come equipped with CTs. Initially, it is crucial to verify the transformation ratio, followed by the nominal burden. Both parameters are vital for the effective management of reactive power.

The minimum acceptable sensitivity of a power factor controller should not be less than 1% of the CT's nominal secondary current. Consequently, this equates to a reactive current of 0.05 A or 0.01 A, corresponding to the CT's nominal secondary currents of 5 A or 1 A, respectively.

Assuming current transformers (CTs) with a ratio of 1000 A to 5 A, which implies a ratio constant \( k \) of 200, are used to measure the primary current of the incoming supply, and the new compensation bank operates in increments of 25 kvar, the resulting \( C/k \) value is 0.083 A reactive at the grid's nominal voltage of 400 V. This value is slightly above the required sensitivity level minimum, ensuring adequate control.

Incidentally, if the step size is, for example, 10 kvar, proper control by the power factor controller is no longer possible. In this case, a separate CT with a lower transforming ratio is to be installed.

With a step size of 10 kvar, a maximum ratio of k = 120 is permissible, indicating the installation of a CT rated at 600 A/5 A.

It is crucial to verify that apparent currents do not exceed approximately 700 A, as current transformers (CTs) are rated to handle 1.2 times their nominal current consistently without experiencing magnetic saturation.

Power losses along CT cables
Figure 2 – Power losses along CT cables (a) for secondary current path 5 A and (b) for 1 A

When ordering a new CT scanner, the following technical data are required:

  1. Type of CT for installing either on cable or at busbar.

  2. Dielectric strength, standardized to either Ur = 0.66 kV or 0.8 kV.

  3. Nominal transforming ratio (e.g. 600 A/5 A).

  4. Rated apparent power (10 VA usually).

  5. Accuracy (Class 1–3).

Concerning the rated power, such as 10 VA, it's crucial to guarantee that the CT is never overloaded. The power consumption of the current path in power factor controllers typically ranges from 0.5 to 1.0 VA. Additionally, if an ammeter is installed, an extra 1.5 VA should be considered.

Table 2 details the typical consumption rates of various measuring devices, which are crucial considerations for selecting the appropriate CT.

Table 2 – Typical values of apparent power, which refer to various measuring devices loading the CT with a 5 A secondary, are provided by individual manufacturers. These values are crucial for ensuring the accurate functioning of the CTs.

Electrical Devices

Consumption per current path


Moving-iron ammeter



Moving-coil ammeter with rectifier



Bimetallic ammeter



Power meter



Power factor meter



Energy meter

AC single phase



AC three phases




Reactive current relay



Overcurrent relay



Bimetallic relay



Power factor controller, electronic



However, the greatest power consumption is attributed to the leads, as outlined in the subsequent calculation.

2.2 Example Calculation of Highest Consumption of Power

Let's compute the energy consumption and power usage in kilowatt-hours (kWh) across different time intervals.

  • Daily Energy Consumption:

    • To calculate the daily energy consumption in kWh, use the following formula:

    • Energy Consumption (in kWh) = Power Rating (in watts) × Usage Time (in hours) / 1000

    • For example, if an 80-watt fan is used for 4 hours daily:

      • Daily power usage in Wh = 80W × 4 Hours = 320 Wh

      • Daily power usage in kWh = 320 Wh / 1000 = 0.32 kWh

  • Monthly Energy Consumption:

    • Multiply the daily kWh by 30 (assuming a 30-day month):

      • Monthly power usage in Wh = 320 Wh × 30 = 9600 Wh

      • Monthly power usage in kWh = 9600 Wh / 1000 = 9.6 kWh

  • Annual Energy Consumption:

    • Multiply the daily kWh by 365 (for a year):

    • Annual power usage in Wh = 320 Wh × 365 = 116800 Wh

    • Annual power usage in kWh = 116800 Wh / 1000 = 116.8 kWh

Keep in mind that 1 kWh (kilowatt-hour) equals 1000 Wh (watt-hours). Therefore, to calculate energy consumption, you should divide the total watt-hours by 1000 to convert them into kilowatt-hours.

  • Length of cable l = 20 m (PFC to CT)

  • Cross-section 4 mm2 (copper)

  • Resistance of the controller’s current path 0.2 Ω

R = 2 × l / κ × A


  • l – Length of the cable (PFC to CT)

  • κ – Reciprocal specific resistance (copper)

  • A – Cross-section

R = 2 × 20 m × mm2 × V / 4 mm2 × 56 A × m = 0.178

The burden of an ammeter can be determined using the formula: Burden (in Ohms) = CT secondary Resistance + Lead Wires resistance + meter/relay resistance.

P = I2 × R

R = P / I2 = 1.5 VA / (5 A)2 = 0.06 Ω

Therefore, the total burden is the sum of the CT cable's burden, the controller's burden, and the ammeter's burden.

Rt = 0.178Ω + 0.2Ω + 0.06Ω = 0.438Ω

At a nominal secondary current of 5 A, the CT would have a burden of: S = I^2 × R = 5^2 A^2 × 0.438 Ω = 10.95 VA, which is approximately 11 VA.

A CT rated at 10 VA would be overburdened. It is essential to either choose a CT with a higher rated power, increase the cross-section of the CT cable to 6 mm2, or consider removing the ammeter.

Then, as previously described, it must be verified whether the ammeter can be connected in series to the current path of one of the main CTs with a ratio of 1000 A to 5 A.

Figures 2a and 2b illustrate the power losses in CT cables based on their respective cross-sections. Figure 2a depicts the losses for CTs with a nominal output current of 5 A, while Figure 2b pertains to CTs with a nominal secondary current of 1 A.

Current transformers (CTs) with a nominal output of 1A facilitate extended transmission distances with comparatively low cross-sections and minimal power losses.

All current transformers (CTs) are labeled with the capital letters 'K' and 'L' on their casings, and their output terminals are marked with the lowercase letters 'k' and 'l'. Therefore, each CT should be installed with the 'K' side facing the incoming supply at the point of common coupling (PCC) and the 'L' side facing the consumers.

As depicted in Figure 1, it is essential for the Current Transformer (CT) to be positioned near the energy meters to maintain the targeted power factor monthly. Typically, power factor controllers measure the electrical load through a single CT, which is commonly installed in phase L1.

However, the controller's voltage path is derived from the other two phases, L2 and L3. Consequently, the vector from L1 to N is electrically shifted by 90° relative to L2 and L3, which allows for the capture of reactive power.

Manufacturers' instructions typically present this configuration as standard. If the current transformer (CT) in phase L1 is unavailable, for instance, in L3, then the voltage path should always be derived from the other two phases, in this case, L1 and L2, and so forth.

Be aware that each power factor controller can attain the predetermined target power factor solely at the location where the current transformer (CT) is installed.

 Connection example Janitza’s power factor controller Prophi
Figure 3 – For connecting Janitza's power factor controller Prophi 12RS, use voltage measurement across L2–L3, 12 relay outputs, target cos(φ) changeover, an alarm output, and an RS485 interface.

It is crucial to ensure that the Current Transformer (CT) also captures the reactive current from the compensation bank. If the CT is mistakenly installed between the compensation bank and the consumers, controlling reactive power becomes unfeasible. This is because the power factor controller would only measure the uncompensated portion of the electrical installation, resulting in the engagement of all capacitor steps without achieving the intended compensation effect from the capacitors.

Turning off the capacitors when finishing becomes unfeasible, even as the current drops to zero, because the current transformer (CT) can no longer measure the leading reactive current from the compensation bank.

In larger electrical plants, energy consumption is typically metered on the MV side. It is crucial to set the target power factor at the controller slightly higher (for example, 0.95 lagging instead of 0.92) because the CT senses on the LV side, and the reactive energy consumed by the power transformers should not be accounted for.

2.3 Preset Switching Time Delay per Capacitor Step

Manufacturers supply compensation banks with preset switching time delays typically ranging from 20 to 30 seconds. Customers are responsible for adjusting these delays to match the anticipated variations in the lagging reactive power demand of their consumers.

If minimal fluctuations are anticipated, switching time delays ranging from 40 to 60 seconds are generally adequate. However, for more frequent variations in reactive power, switching time delays should be between 20 and 40 seconds.

The primary objective of reactive power compensation is to maintain the desired power factor, such as an average of 0.95 over one billing cycle. As a result, it is advisable to review the initial two or three energy bills from the electricity distribution company.

Corrections may be required at any time to adjust the switching time delay or the preset power factor target.

It is not advisable to set switching time delays of less than 20 seconds. Shorter delays can lead to higher switching frequencies. Conversely, longer delays can extend the lifespan of switching components such as air contactors and capacitors.

Additionally, this guarantees that any capacitor step prepared for re-energization will have been discharged since the last period of operation. It is important to emphasize that the suggested switching time delays are applicable solely to LV compensation banks.

In MV compensation banks, various unique considerations come into play. MV capacitors are produced with a rated reactive power starting from above 100 kvar and can reach up to several thousand kvar. The substantial energy stored must be dissipated as heat upon de-energizing a capacitor, regardless of the unloading technology employed.

Medium Voltage Capacitor Banks
Figure 4 – Medium Voltage Capacitor Banks

Classically, this can be achieved using resistances or, more expensively, reactors, and occasionally through voltage transformers.

The discharging devices require cooling, which may take approximately 15 minutes. Consequently, power factor controllers used in medium voltage (MV) compensations should have the ability to preset switching time delays ranging from 15 to 30 minutes or longer.

Reference: Reactive Power Compensation by Wolfgang Hofmann, Jurgen Schlabbach and Wolfgang Just (Purchase hardcover from Amazon)


Rated 0 out of 5 stars.
No ratings yet

Add a rating
  • Beyaz LinkedIn Simge
  • Beyaz Facebook Simge
  • Beyaz Heyecan Simge


bottom of page