PFC in high harmonic environment

[Pafi]

Most of PFC regulators - all that I know - work on cos FI measurement (displacement PF). In a high harmonic content situation, the difference between cos FI and power factor PF is very big and depends on the value of THD. And the Utility Company meter measures PF, not cos FI. Does anybody know a PFC regulator - if it exists - to work on the base of PF measurement - not cos FI measurement? It is almost impossible to filter harmonics to the limit when PF equals cosFI - in my situation.
I am waiting comments.
regards.

 

[busbar]

You may need to investigate filtering for circuits which serve equipment that's causing harmonic currents, and specify harmonic limits for future equipment acquisitions.

 

[1x1x1x]

Pafi,

The power factor (pf) and cos (fi) should be the same, if we talking about the displacement power factor, which is the phase shift between the fundamental current and fundamental voltage. This power factor called pf1 is a result of the controlled firing angle in a thyristor converter application.
In that application the system total power is a vector sum of the fundamental P1; Q1 and the Qh (harmonic) reactive power.
The harmonic components on the line are always producing reactive power, but I am not sure how the power company can measure them. Far as I know, the only concern is the fundamental power factor.
Where the high harmonic distortion is a concern the power factor correction is done using tuned harmonic filters.
The capacitors in the filters are supplying the fundamental reactive power to the system and the filters provide low impedance to the tuned frequency.
With properly designed filters and SVC (in a very dynamic application such as arc furnace), it is probably possible to filter the harmonics below the limit of IEEE 519 and keep the displacement (fundamental) power factor at 0.95.

 

[electricuwe]

If you are talking about a regulator generating switching signals for a capacitor bank for power factor correction you should consider the following:

- The capacitor bank can only control the displacement PF, so it's o.k. if the regulator only evaluates this.

- The regulator should be able to provide a correct displacement PF even if some harmonics (in current and volatge) are present, otherwise it's operation might lead to unnecessary switching of the capacitor bank

- In a high harminic content enviroment capacitors should be used only with a series inductor, usually adjusted to form a series resonant circuit slightly below the fifth harmonic, should be used. Otherwise the capacitors will blow up regardless of the regualtor used.

 

[Pafi]

Thank you for the above answers!
The situation is like that: a single high power consumer - 6 pulses thyristor rectifier(700VDC, 1200ADC rated current)used in a drilling station - for oil - is supplied from a 1600kVA transformer - 20/0.69kV. You can not make a permanent filter for the high values harmonic currents involved because the working regime of the rectifier is not constant - varies very much. On the other hand, a dynamic PFC using steps with series chokes+capacitors connected with tyristors would cost much and - I have to confess - it is beyond my experience. So I chose an automatic bank with 3 steps, the first and second step for 5-th harmonic filtering and the third for 7-th harmonic, using contactors and a regular PFC regulator (KBR from Germany) that has also a program for filters. But surprise came after instalation, the cos FI indication of regulator was different from PF measured with Utility meter - an all digital ABB made. We made our own measurement with a digital apparatus that could measure both cos FI - displacement PF - and PF (with harmonics) and find out that Utility meter measures PF - not cos FI on the fundamental. I believe they do that because in the vast majority of situations cos FI (displacement PF) and PF are equal - with the difference that cos FI can be positive or negative and PF can be only positive. I think that the filtering we made is not very effective and this is the reason for the difference in cos FI and PF. So, even if the regulators work fine on the displacement PF, the Utility meter still counts, measuring PF.

 

[busbar]

As suggested, there are two definitions of power factor. The terse description is that the utility meter most likely uses ratio of watts to voltamperes, as opposed to the more traditional cosine of phase angle between volts and amperes, determined for each phase and averaged. The two methods produce the same results for zero-distortion cases, but the ratio method can produce a lower PF reading where distortion of classically sinusoidal voltage/current waveforms exists.

The first method is frequency independent; the second is not, for it is based on 90° phase shift {id est, timing} tied to a specific line frequency for measurement of reactive quantities. The differing measurement algorithms are discussed in IEEE Std 1459-2000.

 

[Repa]

One of the best products to solve this is Sinewave from Merlin Gerin. Check website

http://www.mgeups.com/download/doc_intl/harmcond/index.htm

I have installed several units reducing THD to 3% and PF to 0.95.

If you have any question let me know.

 

[electricuwe]

Pafi,

correct operation of your system provided you have choosen the most cost effective approach. Your system should be able to improve power factor significantly.

Thy only drawback which comes to mind is a case of dynamic operation (sorry I'm not familar with your application). If speed and torque change very often in your application improvement in power factor will be poor because the controller limits swichting instants to reduce wear on the contactors.

The solution suggested by Repa will be even more costly than a thyristor controlled PFC

 

[Pafi]

electricuwe,

I am a small panel builder for PFC and that application was - and is - my biggest one.
I have just discovered the problem. In fact I knew it and the answers I received here confirmed my supposition. The problem with PF controller is that it measures DPF (displacement PF) in this high harmonic content environment which is very different from PF measured by Utility meter. If the filtering is very effective, than the DPF will almost equal the PF and regulator(and the whole automatic installation) can work properly.
I used iron chokes and capacitors made according to conditions for current and voltage and of course to standards - in terms of value tolerances. And because of this tolerances, the value of the resonance frequency was slightly different from the harmonic value. I thought first that would not be a problem. But I made after that some new reactors, air core, made with very small tolerances and according to the measured value of capacitors, so the resonance value of the series connected capacitor+choke was indeed the one of the harmonic. And I had the surprise to see that the filtering was so effective than PF almost equals DPF and the installation worked very well. I thought that this information must be shared, in the literature you can read that the resonance frequency can be slightly under the harmonic frequency but from my experience and measurements it is better to be as close as possible to the harmonic frequency. One can say that this effectiveness depends on the reactances in the whole system and the relation between them. Still, I think that filter has to be very close to the frequency that we want to filter. Excuse my English!
Pafi

 

[busbar]

For the sake of other readers, it may be appropriate to explain the difference between displacement power factor and “true” power factor. Where there are differences in measurement, true power factor will be lower, for it combines characteristics of traditional phase displacement, but also accounts for harmonic content. Classically the two produce identical results when there is no distortion in the measured voltage and current waveforms.

An important distinction in measurement methods is that displacement pf is based on the Pythagorean relationship—with inescapably frequency-dependent reactive-power quantities. Displacement power factor is defined as the cosine of the angle between the fundamental voltage and current waveforms. True power factor is based on the ratio of real and apparent power at the point of common coupling—typically at the electric utility’s traditional metering point. As such, they inherently account for a wide assortment of voltage and current “recipes.”

To add insult to injury, some electric-rate schedules base billing power-factor calculations on ratios of peak 15/30-minute demand-interval real and appartent power, while others use monthly-totaled ratios of real energy {watthours} and apparent energy {voltamperehours}.

Historically, the IEEE Standard 100 Dictionary definition excluded all but the fundamental power quantities. This limitation is in the process of being modified/updated in 1EEE Std 1459-2000.

[Nowadays, capacitor-input rectifier stages seem to be the greatest overall contributor to harmonic quantities.]

A worthwhile reference is Chapter 10 of T Longhand, et al., Power Capacitor Handbook, Butterworths, 1984. Another is Dr Mack Grady, Harmonics and How They Relate to Power Factor, Proceedings of the EPRI Power Quality Issues & Opportunities Conference, November 1993.

 

[electricuwe]

Pafi,

the approach with filters tuned slightly below the harmonics frequencies is mostly used for general PFC applications whereas for dedicated PFC applications, like in your case, exactly tuned filters might be used.

The background is the following:

in a general PFC application an exactly tuned filter might catch a lot of harmonic currents injected by other customers via the incoming supply and its difficult to avoid overloading the filter by excessive harmonic currents.

in a dedicated PFC application the transformer decouples the PFC from other harmonic sourrces and the harmonics generated by the drive can be calculated quite well so that exactly tuned filters can be designed without unacceptable risk.

 

[jbartos]

Suggestion: The active harmonic filters and hybrid (passive and active) harmonic filters are considered better solutions than the passive harmonic filters.
Visit

http://www.activeharmonicfilters.co.../body_high_performance_active_harmon.html#TOP

http://www.e2i.org/docs/TR-111920.pdf

http://www.isea.rwth-aachen.de/Ww/jahrbuch/2001/forschung/jb_2001-forschung_07.pdf

etc. for more info

 

[electricuwe]

jbartos,

it would be nice if you could name a vendor for an active or hybrid filter in the power range Pafi needs. I'm afraid if such a filter is available it will be more expensive than the thyristor rectifier.

 

[jghrist]

The practice of tuning filters slightly below the harmonic frequency is proper because of the interaction of the filter with the source impedance. The parallel combination of the filter and the source impedance will bave a minimum impedance slightly above the filter tuned frequency.

There is also a resonant point of the parallel combination below the tuned frequency which has a high impedance. By designing the filter with a slightly lower resonant frequency, you reduce the chance that manufacturing tolerance, aging, and temperature will cause a variation that results in the high impedance resonant point being close to the harmonic frequency.