VFD Input Filter

Status
Not open for further replies.
E

Electriman

Senior Member
Location
TX
Greetings,

When we are specifying a VFD, do we need to specify the input line filter or do the manufacturers know what filter to put in? I am assuming all VFD manufacturers have to meet IEEE STD for harmonics injection into grid so they need to have a line filter.
I am also interested to know if these filters are integrated into the VFD package or do they come in their own enclosure?

Thank you in advance.
 
Electriman said:
Greetings,

When we are specifying a VFD, do we need to specify the input line filter or do the manufacturers know what filter to put in? I am assuming all VFD manufacturers have to meet IEEE STD for harmonics injection into grid so they need to have a line filter.
I am also interested to know if these filters are integrated into the VFD package or do they come in their own enclosure?

Thank you in advance.
Click to expand...
Why do you need an input line filter?
 
The really harmonic rich steep edged waveforms are on the output with little leakage back to the input. If the VFD does not have an active power-factor correcting front end the result will be similar to any diode-capacitor rectifier system load. Depending on the rest of the load on the service there may be no need to correct/filter that distortion power factor.
Any IEEE standards for harmonic injection (with which I am not familiar) are not part of the electrical code but may help to satisfy any limitations that POCO puts on your overall load and improve power quality within your premises.
 
All the Vids I have installed have line reactors, is that what you are referring to as a filter? If harmonics, our POCO IS kinda concerned...not enough for the small applications I have done (100 hp and smaller)
 
tom baker said:
All the Vids I have installed have line reactors, is that what you are referring to as a filter? If harmonics, our POCO IS kinda concerned...not enough for the small applications I have done (100 hp and smaller)
Click to expand...
That is what I exactly meant, Input reactor. Sometimes I call it VFD input filter.
 
The VFD manufacturers are not responsible for meeting IEEE519, the installer is. Whether or not you need a line reactor, filter (different things), multi-pulse rectifier with phase shifting transformer or Active Front End VFD to meet IEEE519 is a decision based upon the circumstances of the proposed installation. I could have a 5HP drive as the only harmonic producing load on a 500kVA transformer, and I would meet IEEE519 at the point where the transformer connects without anything in front of it. Conversely I could have 10 x 50HP VFDs on the same transformer, and simple line reactors would not be enough. There are too many variables to make generalizations.

But that's only about harmonics. The OTHER reason, and better one, to add a line reactor ahead of a VFD is to add inductance ahead of it in order to slow down the rise time of transients coming in from the grid. Ringing transients can damage VFDs that are running when they happen, so adding the inductor ahead of them makes the drive less vulnerable. I call it "cheap insurance". But even then, if the transformer feeding the drive has a kVA less than 10x the VFD kVA, then you don't need the line reactor either. Or if you have multiple small VFDs, you can get the same protective effect by having one larger line reactor ahead of the group. So mfrs do not build them in unless asked to. (I once worked for a small mfr that did and they never failed, but they went out of business because their drives were bigger and more expensive than others).
 
Jraef said:
The VFD manufacturers are not responsible for meeting IEEE519, the installer is. Whether or not you need a line reactor, filter (different things), multi-pulse rectifier with phase shifting transformer or Active Front End VFD to meet IEEE519 is a decision based upon the circumstances of the proposed installation. I could have a 5HP drive as the only harmonic producing load on a 500kVA transformer, and I would meet IEEE519 at the point where the transformer connects without anything in front of it. Conversely I could have 10 x 50HP VFDs on the same transformer, and simple line reactors would not be enough. There are too many variables to make generalizations.

But that's only about harmonics. The OTHER reason, and better one, to add a line reactor ahead of a VFD is to add inductance ahead of it in order to slow down the rise time of transients coming in from the grid. Ringing transients can damage VFDs that are running when they happen, so adding the inductor ahead of them makes the drive less vulnerable. I call it "cheap insurance". But even then, if the transformer feeding the drive has a kVA less than 10x the VFD kVA, then you don't need the line reactor either. Or if you have multiple small VFDs, you can get the same protective effect by having one larger line reactor ahead of the group. So mfrs do not build them in unless asked to. (I once worked for a small mfr that did and they never failed, but they went out of business because their drives were bigger and more expensive than others).
Click to expand...

Thanks for your response. My understanding is that we either need a line reactor or filter or nothing. Is there any case that we need to have both line reactor and filter?
Second question, is filter/line reactor part of VFD package or do they come in their own enclosure?
Thanks
 
Electriman said:
Thanks for your response. My understanding is that we either need a line reactor or filter or nothing. Is there any case that we need to have both line reactor and filter?
Click to expand...
A reactor is the first line of defense. A passive harmonic filter is a reactor plus capacitors, tuned to absorb a specific harmonic, followed by another reactor (L-C-L filter). The goal in reducing harmonics, although full of variables, can be thought of as trying to control the Voltage total harmonic distortion (V-THD) by limiting the Current total harmonic distortion (I-THD) which is what non-linear loads like VFDs create. So if you can keep your I-THD from the VFD to 5% or less, you are all but guaranteed to meet IEEE-519 voltage distortion limits at the PCC (Point of Common Coupling).

Quick list of options in order of effectiveness (also generally cost as well):
For reference, a VFD with no reactor, no DC bus choke (DC bus choke is like a reactor but inside of the VFD on the DC bus): 80-120% I-THD, meaning the current flowing in the circuit at harmonic frequencies can actually EXCEED the normal working current of the motor.

Remember, the GOAL is 5% I-THD or less;
  1. VFD with line reactor, OR a DC bus choke: 30-35% I-THD
  2. VFD that has a DC bus choke AND a line reactor ahead of it: 20-25% I-THD
  3. VFD + Passive Harmonic Filter, tuned to the 5th harmonic only: 7-10% I-THD, tuned to the 5th and 7th: 6-8% I=THD
  4. VFD with 12 pulse front end: 7-10% I-THD*
  5. VFD with 18 pulse front end: 5-6% I-THD*
  6. VFD with Active Front End (AFE): <5% I-THD
  7. Active Harmonic Filter (AHF) connected to the same AC input circuit as the VFD (at PCC): <5% I-THD

* Multi-pulse front-end drives require voltage balance between phases to be <1% to meet these values. At >1%, harmonic mitigation diminishes.

Second question, is filter/line reactor part of VFD package or do they come in their own enclosure?
Thanks
Click to expand...
Either way. VFD mfrs (or panel builders) can package the reactor or filter for you, or you can get them stand-alone in their own enclosure and wire it yourself. Multi-pulse and AFE drives come built that way, AHFs are always a separate unit that connects at the PCC.
 
Last edited:
Jraef said:
A reactor is the first line of defense. A passive harmonic filter is a reactor plus capacitors, tuned to absorb a specific harmonic, followed by another reactor (L-C-L filter). The goal in reducing harmonics, although full of variables, can be thought of as trying to control the Voltage total harmonic distortion (V-THD) by limiting the Current total harmonic distortion (I-THD) which is what non-linear loads like VFDs create. So if you can keep your I-THD from the VFD to 5% or less, you are all but guaranteed to meet IEEE-519 voltage distortion limits at the PCC (Point of Common Coupling).

Quick list of options in order of effectiveness (also generally cost as well):
For reference, a VFD with no reactor, no DC bus choke (DC bus choke is like a reactor but inside of the VFD on the DC bus): 80-120% I-THD, meaning the current flowing in the circuit at harmonic frequencies can actually EXCEED the normal working current of the motor.

Remember, the GOAL is 5% I-THD or less;
  1. VFD with line reactor, OR a DC bus choke: 30-35% I-THD
  2. VFD that has a DC bus choke AND a line reactor ahead of it: 20-25% I-THD
  3. VFD + Passive Harmonic Filter, tuned to the 5th harmonic only: 7-10% I-THD, tuned to the 5th and 7th: 6-8% I=THD
  4. VFD with 12 pulse front end: 7-10% I-THD*
  5. VFD with 18 pulse front end: 5-6% I-THD*
  6. VFD with Active Front End (AFE): <5% I-THD
  7. Active Harmonic Filter (AHF) connected to the same AC input circuit as the VFD (at PCC): <5% I-THD

* Multi-pulse front-end drives require voltage balance between phases to be <1% to meet these values. At >1%, harmonic mitigation diminishes.


Either way. VFD mfrs (or panel builders) can package the reactor or filter for you, or you can get them stand-alone in their own enclosure and wire it yourself. Multi-pulse and AFE drives come built that way, AHFs are always a separate unit that connects at the PCC.
Click to expand...
Thank you Jraef! If I could think of a place, I would make this post a sticky.
As it is, I will bookmark it.
 
IEEE 519 establishes harmonic limits is at the utility service, not at the VFD. The VFD manufacturer cannot know how their VFD will mix with all of the other harmonics in the building such as other VFDs, LED lighting, ECM motors, and computers to result in the net harmonics at the Point of Common Coupling (utility service). Even if they pretend to simulate with software, it is an oversimplification.
 
Status
Not open for further replies.
Top