VFD Low Load Power Factor

[heinrock]

Can anyone point me to a reference that can help me estimate power factor of a VFD running a fan load but only slightly loaded? The red book says VFDs have poor power factors at low loads, but that's all the farther they take it.

 

[jraef]

You need to define what kind of power factor you are concerned about. Follow this thread for a reason why (so I don't have to repeat myself and you can benefit from other future comments on it).
thread237-202478

 

[heinrock]

I read that thread and it's good. The thing is that I'm concerned with all power factors since I'm trying to optimize a system for a startup procedure. I need to know how much steam to make so I can spin a turbine to turn a generator to supply enough apparent power to run some blowers by a VFD at some low but presently unknown load. The easy answer is make tons of steam and make the steam drum really big, but those aren't luxuries I have on this project.

The VFD manufacturer, a very large German comapny, advertises, through the technical people that I have access to, and through their manuals, that their new hyfalutin rectifier presents whatever power factor I want it to present to the line using a vector control algorithm. They also say that they keep distortion low. I'd be interested in any unbiased opinions on that one.

Ideally, I wish there was a curve that showed inverter PF over the range of its loading, like some motor manufacturers give you. That way I can generate a realistic worst case for apparent power requirement if I chose to go with slightly older but proven technology (the VFD from the thread), which I'd like to do just for reliability and simplicity reasons. I cannot use the fully loaded blower as the worst case, because at this point in our envisioned startup procedure the blower will not be fully loaded and we're miserly with the power.

The generator winding's ability to withstand the current harmonics is a completely different problem than the one I'm trying to figure out at this point - though it looms in my future too. If I can get them to form wind it then it should be ok.

Regards

 

[jraef]

Well, being that my employer is likely that "very large German company" my opinion would hardly be considered unbiased, but the statements would be true for just about anyone's similar VFD design. I can only assume you are speaking of an "active front-end" VFD. In that case, the claims are true (to a great extent). An active front-end drive essentially is a set of back-to-back inverters instead of a passive recitfier and an inverter. The inverter front-end can then optimize the way the rectifier pulls power from the source so as to minimize harmonics. Eliminate? No. But minimize, definitely. So if harmonics are minimized, the distortion power factor is also minimized.

 

[ScumPunk]

Hi, all,
Unfortunately, I suspect I am another employee of the "very large German company", except if the VFD you are referring to is the one I'm thinking of, I might be able to help clear up a couple of things, having helped to design it..
It's not an Active Front End drive in the classic sense, as the input bridge does not use PWM. However, it's true that the harmonics are less of a problem than with conventional drives, especially at the lower harmonics.
Cheers,
Mort

 

[jraef]

Yes, I suppose that we should have started off asking what size system you were speaking of. Smaller Low Voltage systems use AFE technology, larger Medium Voltage systems use a different method. Same net result however, as far as low pf performance is concerned anyway.

 

[waross]

Am I missing something here?

I need to know how much steam to make so I can spin a turbine to turn a generator to supply enough apparent power

Calculate the real power required. That will have a direct correlation to the amount of steam energy required. You should be concerned with the efficiency of the VFDs and motors, not the PF.
Once you have calculated the amount of steam required to supply the reactive power or VARs of a system, you often never do it again.
You may need a tiny percentage more steam to supply the losses occasioned by the reactive power.
respectfully

 

[davidbeach]

Good point waross, I'd missed that entirely. It is a very good approximation to say that the steam is directly proportional to the real power out. VArs are produced in the excitation system, not from the shaft. It takes a small additional amount of shaft power to produce the real power that goes into the field to produce the VArs but that is so small compared to the system losses it can be ignored. Your steam demand will be directly related to your real power needs of your load and you can forget about VArs (for this purpose) as long as you stay within the capability curve of the generator.

 

[heinrock]

Good stuff guys.

The powers are pretty small - the generator will be 250 kVA but we don't expect a 250 kW load - more like 180 kW at full load.

My question was originally about the power factor of the VFDs, but I think I was asking the wrong question. My question should have been "what is the relationship between mechanical power I need to supply on the input shaft of a generator and known reactive power demand for the loads." Those last couple posts answered this question quite well: very little mechanical power. Surely I need to arrange the electrical circuit for the total apparent power, but the steam power I need to produce will be much closer to the sum of real electrical power delivered (including, of course, losses).

On the point that the VArs get supplied from the excitation system - does this include if the generator is permanent magnet synchronous?

Thank you all very much for your insight.


By the way - no offense to the German company or it's people. Actually I enjoy their control system products very much and meant no ill will toward you guys. It's just that during the sales cycle it's sometimes hard to get all the info you need. Further, during a sales cycle you rarely get to talk to the designers.

 

[davidbeach]

I doubt that your main generator is a permanent magnet machine as it would have no VAr control. Rather what is much more likely would be that you have a small PM generator on the same shaft as the main generator that supplies power to the field. The field needs more power input to produce more VArs and that power comes from the shaft, therefore from your steam. The excitation system uses very little power compared to the generator power rating so the shaft power required to supply VArs is very small.

 

[Marke]

what is the relationship between mechanical power I need to supply on the input shaft of a generator and known reactive power demand for the loads.

The reactive current at the input to the VFD is very low. The quoted Cos(phi) is generally better than 0.95
The issue is as previously described, that the harmonic currents can be significant. This can result in distortion power factors less than 0.7
Drives with no DC bus reactor or AC line reactors have very high peak currents at the input of the drive. These peak currents are very narrow and occur on the crest of the voltage waveform.
The addition of AC line reactors and/or DC Bus reactors will widen the current pulses to a maximum of 120 degrees per half cycle, depending on the value of reactance added.
Active front end technology can reduce the harmonics down to a very low level.
The issues with the harmonic currents are:
Increased losses in the supply (important for smaller generators where the increase in I2R losses can be significant) and interference with AVR operation.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[ScottyUK]

I doin't have much experience on little steam turbines but here are a couple of thoughts from the world of big steam turbines to consider: steam turbines generally do not like zero load, so consider how to maintain a minimum load on the unit. Even worse is a motoring load, so forget regeneration as a means of slowing the fan down.

There will probably be a load rate limiter which will define a maximum rate at which you can increase the load on the turbine. If you try to increase the load beyond this limit the turbine will normally throttle back and slow down if it is island mode. Whether this action would actually shed any load with a rectifier or AFE drive is highly questionable but it might well cause either a turbine trip or the drive to reject the supply if it has sufficient intelligence to monitor it.


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

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

 

[ozmosis]

Just a point on the "big German Company" AFE. The PF can be controlled (set to whatever you require_, either leading or lagging. It just depends on the current rating of the AFE once you know what PF is required, and this would require a few calculations. An AFE is expensive just for PF control, maybe better looking at Active filter.

 

[waross]

In a basic generator in the range of 180 to 250 KVA (or KW) the power for the field will be taken from the main windings and rectified and controlled by the Automatic Voltage Regulator.
An option which gives better motor starting and much less probability of voltage collapse is the Permanent Magnet Generator option. This is a small permanent magnet alternaor fitted to the shaft on the back of the main alternator and exciter. It produces the power for the main field. The output of the PMG is fed to the AVR where it is rectified and controlled.
In both cases the small amount of energy required is supplied by the prime mover.
respectfully

 

[heinrock]

oz, or anyone else who has used the AFE,

what are your thoughts on using the AFE as a power factor correction capacitor for lagging induction motor loads connected across-the-line to the same bus as the AFE, parallel to it?

what I would want to do is modify the AFE's power factor during runtime into a leading one to counter the motor's as the motor starts, analogous to switching on a capacitor.

I'm thinking I would figure out what the reactive current in the parallel motor is, then add this to the requirement for the AFE.

i would do this instead of a capacitor because i have a footprint size issue

rock

 

[electricuwe]

Regarding the operation of the AFE to compensate the power factor of induction motor loads connected in parallel:

If this is an good idea depends on the alternatives you have:

If your supply voltage waveform has low harmonics and you can connect power factor correction capacitors without concern, this will be the more effective option. The power loss of a power factor correction capacitor is much lower than the additional losses you have to expect in the AFE.

If your supply volate is heavily distorted and you would need power factro correction capacitors with choke or harmoncic traps (with the inductances causeing a lot of losses), the use of the available AFE kVA might be an intresting option.