Impact of VFD on Power Factor

[charlie b]

This is a follow up question related to this thread, but with a different focus:
http://forums.mikeholt.com/showthread.php?t=191748

How does the presence of a VFD impact the power factor seen upstream? For example, take a 20 HP, 480V, 3 phase motor, with a full voltage non reversing starter. Assume that it would have a power factor of 85% if this were the only load on the power system. Now insert a VFD in place of the starter. Speaking in general terms (i.e., no calculations needed), what would this do to the power factor as seen by the power system? Would it,

• Remain the same?
• Go to 1.0?
• Increase a little?
• Increase a lot, but not all the way to 1.0?

 

[Russs57]

Won’t get to one but might make .95 with some luck.

 

[charlie b]

Here's why I am asking. I have a copy of the results of a calculation performed on behalf of a contractor. It claims that a facility that has lots of motors, several of which are 75 HP or larger, can operate at or above a .95 power factor, without the need for any power factor correction equipment. That sounds very counter-intuitive to me. I wonder if the use of VFDs for some of the motors is what is pushing the overall PF higher. I also wonder (with a valid reason that I will not discuss) if the contractor is pushing the limits of credibility.

 

[kwired]

Displacement power factor should be nearly 1.0 on the supply side of the drive.

You may have distortion power factor.

Power factor between motor and drive shouldn't change much.

 

[Besoeker]

This is a follow up question related to this thread, but with a different focus:
http://forums.mikeholt.com/showthread.php?t=191748

How does the presence of a VFD impact the power factor seen upstream? For example, take a 20 HP, 480V, 3 phase motor, with a full voltage non reversing starter. Assume that it would have a power factor of 85% if this were the only load on the power system. Now insert a VFD in place of the starter. Speaking in general terms (i.e., no calculations needed), what would this do to the power factor as seen by the power system? Would it,

• Remain the same?
• Go to 1.0?
• Increase a little?
• Increase a lot, but not all the way to 1.0?

It gives you about 0.958PF and it's distortion, not displacement PF.
And don't use PFC capacitors to try to correct it. If you must for other loads on the system use detuned banks.

 

[Jraef]

It gives you about 0.958PF and it's distortion, not displacement PF.
And don't use PFC capacitors to try to correct it. If you must for other loads on the system use detuned banks.

Everything he said. Strictly speaking of displacement PF it would be nearly unity, but the drive creates distortion power factor and the utility measuring equipment can't tell the difference, hence the (roughly ) .95 or better. But pay particular attention to the last part of his statement. Bulk PFC capacitors or APFC systems have been the source of many many VFD catastrophes. If you need to correct PF for other inductive loads, I prefer to use "at-load" capacitors that are switched in only when the motors are running.

Very VERY recently, there are newer "Active Front End (AFE) VFDs that can be operated with a leading power factor (exporting VARs) so that they can be used to help correct other inductive loads on the same system (within a range based on the size). But it's HIGHLY unlikely that they have that yet, it's currently only available at 250HP and up as far as I know. I mention this mostly for posterity, i.e. someone reading this thread in a few more years when this has become more widespread and the issue begins showing up in power system analysis software packages. What I expect to happen is that when you say you have a VFD, the software will ask you if the VFD is AFE or not.

 

[Ingenieur]

for clarity

displacement pf = cos of phase shift between the v and i at the fundamental freq
does not factor in distortion power, ie, any freq other than the fund

distortion pf = P/S

P = the sum of active power for all frequencies
S = the sum of the apparent or total power for all freqs

 

[Russs57]

Charles, I’d say the claims are within reason if the building loads are primarily motors that will have VFD’s added to them. IMHO computer loads are the worst and final numbers are likely to be based on the diversity of the building load.

So able to be achieved but would require very high percentage of motor and resistive loads.

Maybe it is time to share your misgivings about level of credibility? No need to throw anyone under the bus as we are all involved in achieving these goals. FWIW, I’d be skeptical as well. But only you have (perhaps) the additional information to help determine the potential validity of those claims.

For me in my environment I’m more concerned with distortion than PF.