I've got a 600 VAC, 75 HP VFD that I want to place line and load reactors in series with. Usually, I find charts like this that give me a horsepower table to follow as they are the easiest to use. However, today I googled sizing electrical reactors I've found sites have varying information and suggest different sizings.

I've never challenged suppliers on why these are sized the way they are. When I've purchased them it was simply by using the voltage and horsepower ratings of the VFD/motor but now I'm curious as to the rationale behind all these numbers.

With respect to the guidelines linked above and the following reactors: RL-08002, RL-08003.

  1. Why is 3% on the line side and 5% on the load side recommended?
  2. What does the 3% and 5% relate to?
  3. How is this percentage used to calculate the equivalent mH rating (0.4 and 0.7 mH)?
  4. If I wanted to calculate a required reactor size without the use of an equivalent horsepower table, what would be the proper starting point? How can I determine an unbiased value?
  • \$\begingroup\$ Re 4, what bias are you considering? \$\endgroup\$ – Solar Mike Jul 16 at 22:11
  • \$\begingroup\$ @SolarMike, the table shown above gives a value of 0.7 mH for the 5% reactor. Some online calculations I've found say it should be around 0.57 mH for 5%. I'd assume a higher impedance would mean more copper/windings means more money. Not a huge difference but if I'm advising a client to purchase 100 of them I want the most cost-effective option. \$\endgroup\$ – C. Lange Jul 17 at 1:23
  • \$\begingroup\$ I revised my answer to address the above comment, points raised by @J. Raefield and a comment to his answer. \$\endgroup\$ – Charles Cowie Jul 19 at 15:58

Why is 3% on the line side and 5% on the load side recommended?

3% is recommended for the line side because a 3% voltage drop in addition to the "typical" voltage drop that would be seen without the reactors is estimated to be the maximum that is tolerable.

5% is recommended for the load side because that is the maximum that is estimated to be tolerable considering the combination of input line voltage drop, the capability for voltage adjustment by the modulation and the impact of reduced voltage on motor performance.

In both cases, it is assumed that any added impedance will provide significant benefit.

What does the 3% and 5% relate to?

The percentage refers to the voltage drop as a percent of rated voltage at rated current due to the reactive impedance inserted.

How is this percentage used to calculate the equivalent mH rating (0.4 and 0.7 mH)?

Percentage = 100 x current x 2 x Pi x f x L / line-to-neutral voltage

If I wanted to calculate a required reactor size without the use of an equivalent horsepower table, what would be the proper starting point? How can I determine an unbiased value?

You can use the equation above to calculate percentage based on current and voltage. The "bias" involved in the tables are estimates concerning what are the typical distribution system characteristics and what is typically tolerable.

Individual VFD manufacturers may "tune" their recommendations based on the input or DC link reactance built into the VFD, output dv/dt characteristics of the design and their own tolerance data.


Is it good practice to install reactors for every VFD?

It is difficult to determine the need for harmonic current mitigation or the amount of mitigation provided by reactors without a detailed analysis of the facility power distribution system. Such an analysis would estimate the harmonic currents drawn by VFDs and the effect of those harmonics at points of common coupling with sensitive equipment of other utility customers.

Similarly, it is difficult to determine the need for dv/dt filtering at the VFD output and the effectiveness of reactors as dv/dt filters.

Experience may suggest to many engineers that the client’s money is better spent on reactors than on studying a power system that will probably change over time. However that comes down to advising the client where to place a bet.

Installing 18-pulse drives in sizes that are widely available as 6-pulse drives is a better bet but a larger one.

From the standpoint of the individual engineer, the best bet is to do what the supervising engineer says is good practice.

Voltage drop

Although the percent reactance refers to the percent of rated line to neutral voltage at rated current that is the voltage drop across the reactor, the voltage drop seen at the VFD input or at the motor is less because the impedance of the reactors is almost entirely reactive while the effective full-load impedances of the VFD and the motor are almost completely resistive.


Anyone who is dealing with the possibility of harmonic current in power distribution systems should read 519-2014 - IEEE Recommended Practice and Requirements for Harmonic Control in Electric Power Systems. Also consider reading some of the many IEEE papers that discuss that standard and make recommendations related to it.

There are also a lot of IEEE papers dealing with motor insulation stress and partial discharge due to VFD waveforms. Also the related problem of motor bearing currents has been covered extensively by IEEE papers.


Line reactors help protect the VFD, Load Reactors help protect the motor. NEITHER of them can be effectively justified for the sole purpose of "reducing harmonics" in my opinion, because they do NOT reduce the harmonics enough on their own to be of much value. Load Reactors in fact have ZERO relationship to harmonics. If you are concerned about harmonics, a 6 pulse inverter can create as mush as 80% I-THD, and with a 3% Line Reactor that can drop to 30%-35, then a 5% Line reactor drops it to 25% I-THD. But you NEED to be at around 5% I-THD, so no reactor is getting you there.

The best justification for adding Line Reactors is to add inductance to the circuit, in fact the 3% or 5% is the amount of inductance they add (not the voltage drop). Adding inductance creates what is referred to as the "inductive time constant" for any rapid changes in current, because current cannot change in an inductive circuit any faster than that constant (based on the inductance). This then helps reduce the rise time (di/dt) of current anomalies caused by line side transients and protect the VFD rectifier and DC bus from damage. Whether or not you use 3% or 5% Line Reactor is mostly based on the VFD technology involved; small low cost (primarily Asian) VFDs are often built without DC bus chokes, because for them, capacitors are cheaper. North American and EU based drive mfrs, who have to buy capacitors FROM the same Asian companies they compete with, reduce the amount of DC bus capacitance they need by using DC bus chokes. That in turn has similar effects to having a Line Reactor in terms of harmonic mitigation (30-35% I-THD). So if you have a VFD with a DC bus choke, you only need a 3% reactor since it is mainly there just to help protect the drive, and 3% is fine for that. Then a 3% Line Reactor + a DC bus choke gets your I-THD down to about 25%, again, similar to a 5% reactor. If you have a VFD with no DC bus choke, that's when you want a 5% reactor if you want help with harmonics too. But a 5% reactor represents a significant voltage drop through the VFD, meaning that you may end up losing some Full Load Torque capability at the motor. So if motor torque is critical in you application, you do NOT want a 5% reactor.

Why are you wanting load side reactors at all? Do you have some sort of known problem? Load side reactors are there only to help protect the motor from the non-sinusoidal waveform of the output of a VFD, assuming the motor was not designed for operation from an inverter. Other than that, there is no real strong justification for using it anyway, so the 3 or 5% decision is kind of pointless. If you are worried about your motor windings, pay the extra money to upgrade to a "dv/dt filter", which is a Load Reactor plus additional components that actually do help the motor. And if your drive-to-motor distance is in excess of 1000m, upgrade further to what's called a "Sine Wave Filter", but they cost almost as much as the VFD.

  • \$\begingroup\$ In the end application there will be three of these VFDs. I've just learnt it was good practice to install the line reactors for THD reduction. I've seen the 18-pulse converters before but I haven't read on the difference between a standard 6-pulse + reactors and the 18-pulse. Your recommendation on that would be to go the 18-pulse route and forget the line/load reactors? Re: load reactors. Same thing, just something I learnt and I never argued it. Always learning. The motors are about 50 m away. \$\endgroup\$ – C. Lange Jul 17 at 1:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.