Under what conditions is it OK to NOT include a line reactor in an application with a VFD? 3

[rmore]

I need to justify adding the cost of line reactors to a product we are developing that uses a VFD and have a few questions that I'm sure are going to come up.

Note: Our machine requires a dedicated branch circuit.

1. Under what conditions is it suitable to NOT include a line reactor (and it's associated costs)?

2. Since out machine requires a dedicated branch circuit, can the line reactor be placed at the branch circuit panel or does it have to be placed at the machine?

3. Why don't VFD manufacturers build the line reactor into the VFD? Is it simply cost?

 

[Skogsgurra]

1. It all depends. If your drive can be shown to "behave" when measuring conducted emission (different limits, depending where in the world you are) then you can omit the reactor. But if the drive has an AFE or is a four-quadrant drive, you may need to add a line reactor. Especially if the regeneration is via a thyristor bridge. They need it to limit the commutation di/dt. So, there may be a technical reason for a line reactor. Mostly, it is about conducted EMI and then you may even need a filter that stops emission to influence the grid.

2. To be effective, the reactor/filter needs to be close to the VFD. Usually they are put close to the drive. I don't think that I ever have seen it at the PCC.

3. Mostly cost. But also because reactors make the drive heavier and add to the dimensions of the VFD enclosure. There are a few drives that have filters built into them. I think that that is getting more and more common. But mostly, the line reactor or filter is built so they can be placed under the drive.

If you are absolutely positive that EMI limits are not violated under all circumstances, you can ditch the reactor/filter. Talk to the VFD suppliers. And don't believe all they say. Ask for proof. There should be protocols from test houses that show if it is OK or not. The EMC declaration, usually signed by the CEO, is not enough. You shall ask for protocols. ETS or similar test houses can mostly be trusted.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[jraef]

My opinions:

1. Line reactors are "cheap insurance for an expensive box" against line transients that can cause harm to the VFD. But if your source / service transformer is small compared to the VFD, it is going to act in a similar fashion anyway, so the reactor may end up being superfluous (translate; waste of money). The quick-and-dirty rule is that if your source kVA is more than 10x the VFD kVA, then you need a reactor. So for example if your service transformer is 150kVA and you have a 75A 480V drive, the drive is 62kVA out of 150 (source is roughly only 2.5x), so may not need a line reactor. But if the source is from a 750kVA transformer, you need one. If the VFD has an internal DC bus choke, you can use 20x the drive kVA, so in that case you would NOT need one again, unless you are doing harmonic mitigation and want to reduce the harmonics to lessen the burden on whatever other system you are employing (i.e. Active Harmonic Filter).

2. Gunnar covered that.

3. I used to work for a company that brand-labeled a great drive made in New Zealand that had a Line Reactor in every drive whether you wanted it or not. The drives were bullet proof, but cost 10% more than everyone else and they didn't make it... so what does that tell you? Besides, if you have 4 drives on the wall, you can get the same benefit from having one larger line reactor feeding power to them all as you would with 4 separate smaller line reactors, and it will cost you less.


" We are all here on earth to help others; what on earth the others are here for I don't know." -- W. H. Auden

 

[waross]

Thanks for the information Jeff.
The 10 x KVA is a valuable rule of thumb.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[jraef]

I've used that value for almost 20 years now, I have not had a VFD fail due to line transients since implementing that. I've seen the 20x number tossed around a lot by various mfrs, but when I looked into it, they were always in reference to drive product lines that had built-in DC bus chokes, which is why I added that modification.

From what I have observed: VFDs that have been designed and built for and by Asian companies tend to not use DC bus chokes as a matter of course, they are at best an "option". VFDs designed and built for and by EU and North American manufacturers do use DC bus chokes, at least on their high-end products. I attribute this to the simple fact that ALL capacitors are made in Asia and are therefore cheap for Asian manufacturers since they are often made by the same companies making VFDs. The same is true of the transistors too, but that is an unavoidable situation. With caps, EU and NA manufacturers could wind some cheap copper around some cheap steel and reduce the number of caps they have to buy from their competitors, so they went that route. It turned out to have added benefits but I believe, from stories told by "old guys", that the initial decision was purely economic in nature.


" We are all here on earth to help others; what on earth the others are here for I don't know." -- W. H. Auden

 

[electricuwe]

From my point of view Jraef's recommendation is fully reasonable and I can give some more background:

- In case of a lightning surge, the choke will limit peak current and di/dt and hence protect the input rectifier

- In case of a voltage dip, the choke will limit the current when the DC-link is recharged after the voltage comes back