Power factor correction calculation is the process of finding the capacitor power (kVAr) needed to raise a facility's power factor from its current value to a target. The formula is a single line, but with the wrong inputs the panel is either undersized or causes a capacitive penalty. This article explains how reactive power is calculated, the steps of a compensation calculation, how to use the tan φ table and how to size the steps, with a worked example.
The basic formula
Required capacitor power:
Qc = P × (tan φ1 − tan φ2)
- P: the facility's active power (kW). Not an instantaneous value but the typical load the correction targets.
- φ1: current power factor angle; tan φ1 = Q1 ÷ P (current reactive power divided by active power).
- φ2: target power factor angle; because of the penalty limits in Türkiye the target is usually cos φ2 = 0.98–1.00.
tan φ table
| cos φ | tan φ | cos φ | tan φ |
| 0.60 | 1.333 | 0.90 | 0.484 |
| 0.65 | 1.169 | 0.92 | 0.426 |
| 0.70 | 1.020 | 0.95 | 0.329 |
| 0.75 | 0.882 | 0.97 | 0.251 |
| 0.80 | 0.750 | 0.98 | 0.203 |
| 0.85 | 0.620 | 1.00 | 0.000 |
Practical rule: Qc = P × k, where k = tan φ1 − tan φ2 is the difference between two table values.
Worked example
- Active power P = 100 kW, current cos φ1 = 0.75 → tan φ1 = 0.882
- Target cos φ2 = 0.98 → tan φ2 = 0.203
- Qc = 100 × (0.882 − 0.203) = 100 × 0.679 ≈ 68 kVAr
- Current reactive power Q1 = 100 × 0.882 = 88 kVAr; after correction Q2 = 100 × 0.203 = 20 kVAr
- Apparent power falls from 133 kVA to 102 kVA; about 31 kVA of capacity is freed on the same transformer
How to calculate reactive power (finding the current state)
To know φ1 for the formula you must measure the facility's present reactive power. There are three ways:
- From the bill: monthly inductive reactive energy (kVArh) ÷ active energy (kWh) = average tan φ1. For example 44,000 kVArh ÷ 50,000 kWh = 0.88 → cos φ1 ≈ 0.75.
- From the meter: the ratio of the 5.8.0 (inductive) and 1.8.0 (active) index differences gives the same result.
- From an analyser: read instantaneous kW, kVAr and cos φ; log at least one week and evaluate day, night and weekend profiles separately.
Step sizing and practical rules
- Split the total kVAr into 6–12 steps with the smallest step at 2.5–5 kVAr so the controller can fine-tune.
- Choose the total about 10–20% above the calculated value; capacitor capacitance declines over the years.
- If night load is very low (e.g. 10 kW), even the smallest step may cause capacitive overshoot; a shunt reactor or finer steps are needed.
- In drive-, UPS- and LED-heavy facilities use detuned (filter-reactor) capacitors; otherwise even a correct calculation ends with early capacitor failure.
- For fixed correction at individual motors, the capacitor must not exceed 90% of the motor's no-load reactive power.
Verifying the calculation with live data
A correction calculation is done once and forgotten, yet the load profile changes, capacitors age and the ratio creeps toward the limit. The 6S Teknoloji platform calculates the inductive and capacitive ratios hourly from the main meter and the reactive power controller, shows the remaining margin to the penalty threshold within the month and flags the hours when panel capacity falls short. Use the reactive penalty calculator for a quick estimate and see the reactive penalty prevention and power factor correction monitoring solution pages for continuous tracking.
What is the power factor correction formula?
Qc = P × (tan φ1 − tan φ2). P is active power (kW), φ1 the current and φ2 the target power factor angle. tan values come from the table; e.g. raising a 100 kW load from 0.75 to 0.98 needs 100 × (0.882 − 0.203) ≈ 68 kVAr.
How do I calculate current reactive power from the bill?
Divide monthly inductive reactive energy (kVArh) by active energy (kWh); the result is the average tan φ1. 44,000 kVArh ÷ 50,000 kWh = 0.88 → cos φ ≈ 0.75. The same ratio comes from the 5.8.0 and 1.8.0 index differences on the meter.
Should the panel be larger than the calculated kVAr?
Yes, allow about 10–20% margin because capacitor capacitance declines and load may grow. But the smallest step must suit the night load, otherwise a capacitive overshoot penalty results.