Jump to content


Photo

Information on varying the cycles


  • Please log in to reply
6 replies to this topic

#1 Richy

Richy

    Junior Member

  • Members
  • PipPip
  • 3 posts

Posted 20 September 2002 - 02:44 AM

My name is Richy and I have just joined.I
attend a poly Tech in New Zealand.I have starting research in power factor correction and have come up against a brick wall.Any suggestions at all will be of a great help
-I have being trying to find information on having just one capacitor bank.The capacitors are in parrallel with the load. The thyristor will switch the whole bank in each time and through varying the time the capacitors are on will control the power factor.Any explanation,book or web site will be helpfull. I have searched 100's of web sites and just maybe I could of missed one.

- If anyone is still reading it up to this point.If the bank is say 100 uF and I switched it in for 1 cycle every 10 then then could this be seen as 10 uF for 10 cycles.

Any suggestions would be appreciated
Thank you

#2 BigMax

BigMax

    Intermediate Member

  • Full Member
  • PipPipPip
  • 43 posts
  • Location:Melbourne, Australia

Posted 20 September 2002 - 03:36 AM

Welcome Richy!

Hmmmm, your concept sound interesting, though I'm not so sure it would achieve the desired result in such a format......

Using your example case, assuming that 10uF would produce close to unity power factor, maybe at first glance the ratio of connecting a 100uF capacitor for one cyle in 10 would also produce (an average of) unity power factor, however I suggest that the result would be less than desirable.

To my way of thinking™, even assuming perfect switching at the zero crossing, the result would be large overcompensation for 1 cycle, resulting in excessive capacitive Kvar/I, followed by 9 cycles of lesser undercompensation, resulting in inductive Kvar/I.

As the desired result of power factor correction is a displacement power factor of as close to unity as possible, the 'PWM' system you describe would seem to provide something else........, or would it?

I await for the opinions of others.

Cheers!

BigMax:D:D

#3 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 21 September 2002 - 02:16 AM

Hello Richy


QUOTE

If the bank is say 100 uF and I switched it in for 1 cycle every 10 then then could this be seen as 10 uF for 10 cycles.


It is certainly possible to vary the average of many systems by cyling from a high level to a low level, but this is only useful where the conditions are such that the system itself has enough inertia to smooth the variations out. For example, you can vary the output from a heating element by switching between full power and no power provided you do this faster than the time taken for the environment to notice a change in temperature. We can do this with light, provided that we are switching at greater than 50Hz due to the persistance of the eyes.
In the case of the heating system, if we want half power, we switch to full power for half the time and no power for half the time, providing an average of half full power.

Looking at the power factor correction scenario, firstly, we are adding power factor correction in order to minimise the supply losses due to i2R. If we can minimise the current, we will minimise the i2R losses. Most electrical systems have two components of current flowing, resistive (or real) and reactive current (typically inductive). The resistive current provides the real power for the driven load nd the inductive current magnetises the core, or creates the magnetic flux in the system or load. We can reduce the inductive component by adding capacitive current, however if we add too much capacitive current, the reactive current will climb again i.e. it will change from and inductive or laging current to a capacitive or leading current. The reactive current will be zero when the capacitive current is equal to the inductive current.

With your proposal, you will provide excess capacitive current for a period of time, and little or no capacitive current for the remainder in such a fasion that the average capacitive current will equal the inductive current. If we take this scenario, and the switching cycle is short enough, then the average power factor as measured by monitoring cos phi, would indicate a unity power factor. This would appear to be an advantage, however, if you monitor the current, you will find that the current will increase rather than reduce, or may remain the same. More importantly, we should explore what is happening to the supply losses, and this is where we will definitely be in trouble. If you have a capacitive current of twice the inductive current, and it is on for 50% of the time, then the supply current and losses will remain the same. If the capacitive current is greater than twice the inductive current, then the average current and losses will climb which defeats the purpose of power factor correction.

I believe that this idea in it's simple approach is not an option for improving the true power factor, however it will improve the average COS phi which achieves nothing except fooling some meters. Most tarrifs are based on true power factor which is the ratio between KVA and KW and the KVA Maximum demand is frequently measured.

Best regards,

#4 Richy

Richy

    Junior Member

  • Members
  • PipPip
  • 3 posts

Posted 23 September 2002 - 01:20 AM

Thankyou BigMax and Marke for your comments. Due to reading them I feel a bit embarrased for even contemplating that idea. The logic that you provided is very sound.

The first part of my correspondance mentioned switching the whole bank of capacitors in and by varying a portion of each cycle to suit varying levels of reactances.The hard part is turning thyristors as in scr's off while current is flowing through them. I have contemplated using back to back IGBT's.
Has anyone an opinion on this idea.

Thankyou

#5 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 23 September 2002 - 01:33 AM

Hello Richy

I think that the same argument still applies. Remeber that Power factor is the ratio of KW to KVA. By PWM modulating the capacitors on the circuit, you would be creating high harmonic or switched capacitive currents on the supply and that would be creating a poor power factor due to current waveform distortions.
There is a method that you can apply IGBts etc to improve the power factor and it will also reduce the harmonics in the supply voltage. You essentially use an inverter to add/subtract to /from the voltage waveform. You can use this to effectively add a leading sinusoidal current also. Note you need some form of low pass filter to prevent the PWM carrier from being super imposed on the supply.

Keep an eye on the current waveform!!

Best regards,

#6 Richy

Richy

    Junior Member

  • Members
  • PipPip
  • 3 posts

Posted 03 October 2002 - 10:48 PM

Thankyou Mark,Your comment was correct.Could I ask 2 additional questions


-When a capacitor is placed in a shunt configuration, and the thyristor which controlls the capacitor for some reason has been turned on with the supply voltage been at 400V RMS when first switched on.I believe that the supply voltage will see the capacitor as a short circuit with only some thyristor and the capacitors own small series resistance.I would not want to add a large series resistor due to power wastage and an added delay in charging up the capacitor.I would like to add a inductor to slow the rate of the current eg limit the current from an inrush current.Could you help me out with a formula in obtaining the value needed of inductance to do this.

- I read in books about harmonics and don't quite understand them.Could you put me right with my nderstanding. When something is turned off quickly there will be high frequencys involved in producing that straight line eg fourier.To calculate power is it only the current of the same frequency as the voltage I take into consideration and the rest placed in the reactive basket or not.

Any correspondance will help in being put on right track in understanding something that seems to have a new twist every time I turn the corner

ThankYou

#7 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 04 October 2002 - 12:31 AM

Hello Richy


QUOTE

When a capacitor is placed in a shunt configuration, and the thyristor which controlls the capacitor for some reason has been turned on with the supply voltage been at 400V RMS when first switched on

Yes you are correct, the capacitor will look like a short circuit and cause the SCR to fail. The addition of an inductor in series will restrict the di/dt and could be selected such that the maximum di/dt is less than the maximum rate of rise of current of the SCR, however, this will not restrict the current through the device and you would need to ensure that the short term current rating of the device is high enough to withstand the current peak. If you reduce the rate of rise of current enough, then you can reduce the maximum current within safe limits. One problem with putting an inductor in series with the capacitor and then switching at other than zero voltage, is that the inductor and capacitor become a series resonant circuit which you are exciting with a transient and so ringing currents will occur. These can be dampened by the addition of resistance to drop the circuit Q.

The addition of inductance in series will also reduce the effective capacitance and therefore the correction, but provided that the inductive reactance is much higher than the capacitive reactance at line frequency, this will not be a major problem.
The addition of "detuning" reactors to power factor correction circuits is becomming more common as a means of reducing the harmonic currents. Typically these form a tuned circuit with the capacitors with a resonant frequency between 100 and 200 Hz. (189Hz is common)

QUOTE

When something is turned off quickly there will be high frequencys involved in producing that straight line eg fourier.To calculate power is it only the current of the same frequency as the voltage I take into consideration and the rest placed in the reactive basket or not.

Not quite correct. Any waveform that is not sinusoidal can be shown to be made up of two or more sinewaves of different frequencies and amplitudes. If the waveform is periodic, then there will be many sinewave at frequencies that are multiples of the fundemental. These are called harmonics. Thus, if we take a sinewave and distort it by using a rectifier or similar, we are creating lots of harmonics. To calculate the power, take the instantaneous product of voltage and current and sum this. The greater the sample rate, the beter the accuracy. The harmonics definitely can add to the power consumed and can not be discarded!

Best regards,




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users