VFD's and Total Harmonic Distortion. 3

[Chuk1t]

I have a client who wants to keep the Total Harmonic Distortion for current less than 5% at the VFD breaker. What is a reasonable value of % current THD that would be expected for a 12 pulse drive? Supply voltage is 480 Volts.

Motor sizes range from 10 to 50 HP

 

[davidbeach]

What do the drive specifications you have evaluated give as a current THD level?

 

[dpc]

A 12 pulse drive for a 10 hp motor sounds like serious overkill.

 

[jraef]

SERIOUS overkill!
Probably could have met it with a 6 pulse drive and a reactor if it was to measured only at the VFD breaker, because that implies no other contributing factors need to be considered.

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[ozmosis]

10-50Hp seem quite low rating to be looking at 12-pulse drives. Are they connected to a real weak supply?
I have come across tender specifications requesting 5% THID before and when I investigated the reason behind it, some consultants have confused THVD with THID and others have been 'strongly' influenced by companies offering active/passive harmonic filters to reduce THID to these very low levels. The first question to ask is what is necessary. Of course, having low levels of THD is best, but at what cost? fitting 12-pulse drives or various types of filters will typically double (at least) the overall cost of standard drives.

 

[DickDV]

The answer to your question regarding total current harmonics is not simply determined by the VFD. The power network characteristics, transformer impedance, lead lengths, other loads, and VFD design all impact the resulting harmonics.

Many drive manufacturers have computer-based programs which will predict harmonic levels with remarkable accuracy when you input the above network characteristics. Some of these manufacturers will run the analysis without cost to you.

I would have this analysis done before adding anything in the way of extra cost to the system. As has been noted already, you may already be on the verge of overspecifying a solution. I agree with the earlier posters who suspect that a standard six pulse drive possibly with input reactors will meet your needs.

 

[Chuk1t]

Thanks for all the input. I agree that 12 pulse is overkill. I agree with DickDV in that an overall network analysis should be done however the client is focused on only the VFD contribution in isolation.

 

[dpc]

IEEE 519 has a limit of 5% THD at the point of common coupling with the utility. But this is voltage distortion. Currrent distortion can be much higher, and usually is.

There is no practical benefit to limiting current distortion at the drive terminals to such low levels.

Drive suppliers should be able to provide harmonic signature data for their drives at various load points. The harmonics will vary with the drive output. They may not have data for a 10 hp 12 pulse drive, but as a percentage, measured at the drive, 12-pulse drives should all behave in a similar manner.

 

[DickDV]

Just a footnote to dpc's post above. Current distortion is usually worst at light drive loads and gets better as the drive output loading increases. Some manufacturers have taken advantage of this feature to use a saturable (or sometimes called a "swinging") DC bus choke. This device increases inductance as current decreases and decreases inductance as current increases thus allowing a rather large inductance when distortion would be at its worst and a smaller inductance at high loads to avoid low voltage problems.

But, as dpc says, to focus on current distortion at the drive is highly unusual and probably not particularly useful when viewed from a network perspective.

 

[LionelHutz]

I will say that it is unlikely you will reduce the THID to below 10% using 12-pulse drives. In practice, you will likely end up somewhere between 10% and 25% THID.

If the client is using IEEE519 and wants 5% THID this means his short circuit current divided by his load current ratio is <20 and it's a fairly weak system.

However, if the client wants his site to meet IEEE519 and has other linear loads at the site then he's interpreting the standard wrong. The measurement is typically done at the utility connection point. Say the system has 1000A total load and 100A of that load is non-linear. Then, roughly, the THID distortion the non-linear loads can create would be 5% * 1000 / 100 = 50% while still maintaining the 5% THID at the utility entrance. Interpreting the standard correctly means that 50% THID could easily be met with a line reactor and would save a lot of money compared to using 12-pulse drives.

 

[dpc]

DickDV,

Thanks for the info on the saturable choke - that's interesting.

 

[ozmosis]

DickDV
Just a note to your footnote: % current distortion is worse at lower loads. So, although the % THID is higher, the actual current distortion is still relatively low as the actual current drops. In Hong Kong a number of the consultants there stipulate % Current distortion levels at any speed should be no more than 30%. crazy. It wastes our time trying to gently explain why using % current is not the best method to impose harmonic levels and theirs too.

 

[Marke]

Although the voltage THD may be down to 5% and lower depending on the supply network caharcteristics, from my experience, the currnt distortion will be well outside this spec on a 6 pulse system. I would suggest that 5% voltage THD is achievable, but 5% current THd is a tall order.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[Marke]

I am concerned about the advise given in regard to the THD of the current on the input to a VSD.
Al the literature and measurments that I have seen suggest that there is a current THD of minimum 30% when appropriate DC reactors or AC reactors are used when a six pulse rectifier is employed. This is well above the required 5%.
See figure 2 of

http://www.mirusinternational.com/downloads/Power Quality - A New Solution for Harmonics.pdf

and

http://www.unicous.com/oilgas/solutions015.html

Am I missing something here??

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[LionelHutz]

No Mark, You're not missing anything. Right around 30% current distortion is quite normal for a 6-pulse rectifier with a line reactor or DC choke. As you already posted, meeting 5% current distortion at the feed to the drive is a tall order - a really tall order.

 

[ozmosis]

I would say even higher with a line reactor/DC choke: more like 35~40% current distortion. We have a special HVAC drive with the DC link capacitors (plastic film) cut right down to about 5% of a 'traditional'(with electrolytic caps) 6-pulse drive and this, at best, is 28%. So, I agree, getting any lower than this will cost many $/£'s in filtering/active front-end.

 

[ozmosis]

Mark
The original enquiry was about using a 12-pulse drive, and the subsequent THID. I think everyone has been fairly consistent with their advice in saying a 12-pulse solution is going to be very expensive, the need to get down to 5%THID should be questioned and any other solution to achieve this level of THID will inevitably drive cost very high. With all due respect, what is the concern?

 

[Marke]

Hi sed2developer

The original poster has a specification of less than 5% THD current, and questions whether the 12 pulse would achieve that value.
Reading through the thread and responses, I feel that it could be interpreted that a 6 pulse would actually achieve that figure which it will not.
I agree that in a drive of that size, it would be unusual to demand a 5% current THD and that a 5% voltage THD would be a more normal specification and that this would be achievable with a 6 pulse drive of this size.
I am just concerned that the poster could interpret the thread that 12 pulse would not be necessary to achieve the quoted specification rather than questioning the spec itself.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[ozmosis]

Mark
Understood.
I agree with your sentiments that focus on "why 5% THID is requested" should be the main issue.

 

[gonso]

I agree with all previous statements questioning the 5% THID spec. Neither 6, nor 12 drives could do the job. I have never encountered 12-pulse drives of this low rating.

I am missing comments on "active front-end" drives. I've worked with one of these - rated at 250kW. It had line reactors on the input and output sides. The THID was about 2%, at any load-point, from motoring to generating (maintaining an input power factor near 1!). I haven't seen these drives at lower ratings, however.