Author Options:

CFL power factor and how to correct? Answered

I have installed solar pv on my house and using a killawatt meter I have been checking the power factor and they are pretty bad .7.  Residential metering dont charge for power factor but when the solar pv is operating I believe the losses are coming from the pv power generated. Does anyone know of a CFL with a good power factor? I have heard that the power factor for the CFL's could be corrected with a .50 cent part. I believe that would be a capacitor but what size and how could you add it?   ?



Best Answer 10 years ago

Be very careful to find a HIGH voltage capacitor to try this with. Look for good quality, class X dielectric capacitors.

Without access to the right test equipment, about all you can do it experiment by adding parallel capacitance and look at the readings on your meter.

Start with a 0.1uF.



7 years ago

Power Factor losses from electronically ballasted fluorescent lamps are in the form of a "non linear load". It's the action of any switch mode power supply that switches on and off 120 times per second at or near the peak of the ac sine wave to top up the input filter capacitor. This power factor loss isn't the same as a traditional ballast - so it can't be corrected in the same manner. Power factor correction in switch mode power supplies has to be designed into the circuit by the manufacturer, so it isn't like the old days where a simple capacitor would shift the current + volt waveform back in line.

This may sound bad - but in fact it's you're lucky day :) Since you're using solar, and not mains electricity, this power factor loss actually doesn't exist here. It's because batteries and solar, or whatever happens to be naturally DC - these systems aren't affected by non-linear loads like the mains is. The DC systems can supply current and stop just as fast as the switch mode psu demands, so there is no loss from non-linear loads. The inverter is simply a set of semiconductor switches, or a switch mode psu itself that reverses polarity rapidly to closely simulate the mains waveform, but is more or less a rapid polarity flip-flopper to the source DC current. So the inverter too, is not affected by these losses. The Killawatt, and any meter is designed for mains use is actually measuring something that doesn't exist on your solar/battery system. About the only real thing you have to worry about if you had a transformer of sorts of the traditional type in your system in which cfl or other switch mode psu's were secondary too, then the transformer would be appropriately de-rated for the VA's.


9 years ago

Don't know if you are still following this thread but here is your answer:

I am an energy engineer and have been using fluorescent lamps in table lamps for 20 years or so. I've also had permanent energy metering equipment installed on my house for the last 12 years. I've also used Fluke power meters for measuring. Both types of metering measure kW, kVA, and power factor. Specifically I had known for years that CFL's had low power factors (0.5 or so) that the manufacturers are also aware of this fact. They make ballasts for larger fluorescent lamps such as 4 and 8 foot lamps that have had high power factors for years. The mabufacturers answer as to why they didn't have high power factor CFL's was because of their small size. Obviously if government had mandated it they would have made high power factor cfl's.

A few years ago I decided on a power factor correction experiment just to prove that manufacturers can do it. I built a power factor box with variable capacitance that had electric plugs mounted and the capacitance could be turned on and off. That way I could take an appliance or light and place the capacitance box between the wall outlet and the appliance or table lamp. My metering equipment was set up to instantly measure amps, volts, power factor, and wattage.

I was only able to correct the power factor with a capacitor on compact ballasts that had a magnetic (steel transformer as a major component) as opposed to electronic ballasts. On electronic ballasts the power factor dropped slightly rather than going up. So you should have the equipment to measure for power factor before making the changes. With my capacitor box I could verify it worked before mounting a capacitor on the lamp ballast itself.

On a Lights of America ballast for w 22w circline lamp FC8T9 (used to sell them at Wal Mart a couple of years ago and are made in the USA not Chinese junk that’s common now) the addition of a 5uf capacitor had the following effect:

Before adding capacitor:
Watts = 20 , Power factor = 0.556

After adding 5uf capacitor (in parallel across incoming voltage leads):
Watts = 19, Power factor = 0.98 where wattage = volts x amps x Power factor

On a GE ballast with a 13W PL type lamp (stick type also sold by Phillips):

Watts = 14, power factor = 0.587

After adding a 5uf capacitor:
Watts=14, power factor = 0.9

I believe a smaller rated capacitor would have given a higher power factor on the GE ballast w 13w as a 7.5uf capacitor had a P.F of about the same as with no capacitor.

You have to make sure the capacitor is disconnected from the wiring system by a switch when not in use which is why the capacitor for the lamps should be soldered into the actual ballast. My 5 uf capacitors are about 3/8’ x 1” x 1.25” in size for the lamp ballasts. When used with the L of America ballasts the power factor is about 0.98.

Adding these capacitors will reduce amperage by about ½ but the real power will be the same and the electric bill will show no reduction as you are billed based on watts not volts times amps (apparent power). Only in one case could I reduce actual power by adding capacitors (magnetic ballast for a 22 and 32 watt circular fixture where I added a 12.5uf capacitor watts went from 60w to 51 watts and p.f for 0.58 to 0.95).

I was able to correct the power factor permanently on my refrigerator from 0.92 to 0.97 with a 7.5uf capacitor On the dishwasher a 70uf capacitor bank worked while the washing machine was over 100uf.

I just did it to prove it could be done. Been doing this for years with no reliability problems.

Edward D


10 years ago

When you hooked this up did you use a proper "Transfer Switch"?  If not then you are likely to kill someone working on the power lines.


Answer 10 years ago

 I have installed my pv system with proper permits. It will shutdown if utility goes down it parallels with utility