VFD Harmonics 1

[X49]

I'm designing a facility with a 600V, 1000A electrical system and the majority of the loads are 6-pulse VFD's with input reactors. The utility has informed us that we do not have to meet the IEEE 519 standards for harmonics.

Would adding harmonic filters to the VFD's still be worth the expense in this case? What possible harmonic-related problems could occur within the facility if they're not installed? (The primary loads are large VFD's, small motors and building loads, UPS for PLC control system.)

 

[electricuwe]

You will see a voltage waveform that is closer to being trapezoidal than sinusoidal. Usually this does not cause significant problems.

If your larger VFD are having 12-pulse input rectfifiers the harmonics from the 6-pulse units may cause imbalance between bridges that are connected in parallel.

 

[MAGTiger]

That's a pretty general question. It really depends on the specifics. What % of the total load is VFD? What is the total load? What is the source impedance? Are there any voltage changing transformers in the facility? What kind of lighting is ther and how much of the load is lighting? What is the size breakdown of the VFD's? Do you know if the VFD's have internal magnetics? What does the load profile look like for the larger VFD's? Where is the nearest PF bank? How far are you from the substation. ...

These questions need to be addressed for a quality answer, but for a quick and DIRTY answer. If your source is stiff (low impedance), then you are probably ok. If your source is soft, high impedance, then the voltage may get ugly. The VFD's probably won't care, but some lighting and control systems might not play well.

If you are just now in the design phase then I would definitely pursue a quality answer. You'll save yourself headaches in the future as the system grows.

Neil

 

[MAGTiger]

Sorry, I just saw that you have input reactors, that's a big help. But i'd still seek professional help because it will be cheaper to prevent problems than fix them later. Design phase is a great time to get it right.

Neil

 

[X49]

Thanks MAGTiger. Here are some specifics:

There are two 250hp Schneider 6-Pulse VFD's. All other plant loads are less than 50kVA total, and mainly supplied by a 120/208V three phase transformer. Lighting is all fluorescent. The lighting transformer could be harmonic mitigating or k-rated if this helps. The main utility service transformer will be 750kVA, so I suspect the impedance is fairly low for the existing loads.

In the future, up to two additional 250hp VFD's will be installed to meet the process demands.

I suspect that harmonics would not be an issue until the new loads are added many years from now. Would it be alright to allow for space for harmonic filters to be installed in the future but not worry about them for now?

 

[ScottyUK]

You may be better off in the long term if you pay a little more for drives with an active front end, the cost of which you can offset against the provision of a couple of extra circuit breakers in your panel. I don't know Schneider's US products well enough to say whether they have that option, but other manufacturers certainly do.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[DickDV]

If you have six-pulse drives that run approximately at the same time and you can divide the kw about equally, you can get the benefits of 12-pulse drives without the expense by buying a transformer with a wye and delta secondary and powering half of the drives on one secondary and half on the other.

You could even divide the load in thirds and buy a transformer with three secondaries, a wye, a delta, and a zigzag. These are intended for 18 pulse drives but if the loads all operate together, put one third of the kw on each secondary and you will have 18 pulse performance.

 

[LionelHutz]

I'd just like to know how you expect to run 4 x 250hp motors plus 50kVA of auxilary load off a 750kVA transformer or why you would believe a 750kVA transformer at 600VAC makes a 1000A service.

Line reactors will certainly help with the harmonics, likely reducing the current distortion from 100% plus levels to around 30% or so. The solution DickDV suggests would do even better, likely getting below 15% current distortion. Any other "off the shelf" product to reduce the harmonics further gets quite pricy. A single custom passive harmonic filter or a single active filter would likely be price competitive when looking at 1000hp of VFD to filter.

 

[MAGTiger]

Putting the 2 drives on a phase shiftimt 12 pulse xfmr is worthwhile so long as you can afford the xfmr, they aren;y cheap. And so long as the drives have similar load profiles. Unless that 750kva xfmr is K rated, you are going to be approaching it;s capacity with the 500hp and 50kva of additional. Considering you need extra capacity to handle the harmoics those drives will issue. I'd get a cost comparison of the 12 pulse xfmr vs. 2- 3% line reactors. Then go with the cheaper solution for now. Leave room for harmonic filters for the "added later" drives and evaluate the xfmr capacity and loading at that time. But I agree with LionelHutz that the 750KVA isn't going to be adequate in the long run.

Neil

 

[ozmosis]

X49
You indicate the utility company is advising your harmonic distortion will be too much: by how much and on what basis have they made this assumption?
It is quite feasible to introduce either active or passive harmonic filtration at a later date and as long as you have the space and ability to connect, it may be prudent to do it this way as the simple answer is that actual measurement will define the level of harmonic mitigation you need.
Once you know this, sometimes you will find that the level of harmonic canceling is minimal. 12 pulse configurations need to be rated for the full amps of the drive system whereas active filtering only need to be rated for the offending harmonic amps.