Power Factor correction with CFL bulb question?

icarus · July 2009

Here is an strange question,

If you are off grid, and you have a PF corrected device where is the power loss.

For example,, if a CFL burns say 50 watts,, but has a PF of .5,, is it actually drawing 100 watts out of the inverter/battery?

Tony

BB. · July 2009

Re: Power Factor correction with CFL bulb question?

No, the CFL still draws 50 watts and the Inverter is still supplying ~50 watts...

However, it is 50 watts + a little bit more:

P=V*I
P=I^2*R

P=V*I; I=P/V= 50W / 120V = 0.42 amps
P=V*I*PF; I= P/(V*PF)= 50 W / (120 V * 0.5) = 0.83 amps

Ratio of increased power loss = P high PF / P low PF = (0.83a)^2 / (0.42a)^2 = 4:1

So--the I^2*R resistive losses are 4x higher because of the high power factor (as well as limitations of transformer saturation current, etc.)...

If you look at inverters, the VA rating for many is usually higher than the Watt ratings (guess which number Marketing puts on the sales literature)--so they can support, to a degree, the larger PF load without actually consuming too much more power from the battery.

The Outback inverter product line is actually rated in VA and not Watts--so in their case, the PF will ultimately define how much "real work" you can get out of their devices.

-Bill

icarus · July 2009

Re: Power Factor correction with CFL bulb question?

Thanks Bill,

Like I have always said, there are some very smart folks here! I suspect I could pose the question,,"How high is up?" and in a few minutes get a learned response,

Tony

RCinFLA · July 2009

Re: Power Factor correction with CFL bulb question?

In the case of CFL, the design centers around how the high voltage DC is generated.

Cheap way is to just rectify the 120 vac to produce about 160 vdc.

The 160 vdc is chopped at some higher frequency, typically 25 KHz, then fed into a conventional inductive ballast in series with the bulb. Because of the high frequency, the ballast is wound on a ferrite core and is much smaller then the old 60 Hz inductive ballasts.

The simple rectification with a moderate filter capacitor results in most of the input current being drawn during the peak of the input sinewave voltage. The rest of the AC cycle draws no current. This translates to non-sineusoidal load current from the AC main, and therefore a reduced power factor. It is not the same as an electric motor that is usually inductive load dominate which has a sineusoidal load current but is slightly out of phase with the voltage waveform.

A true RMS current meter is required to get true rms input current on this type of CFL since current draw is not a sinewave.

In Europe there are tighter regulations on power factor from devices that draw relatively low wattage like a CFL. In this case the high voltage DC voltage generation has a separate switching circuit that creates the high voltage DC by pumping at a high frequency that follows an average sinewave current profile from the AC mains. This extra circuitry more then doubles the electronics in these PF corrected bulbs.

As stated by BB., because of the higher peak current to achieve the same actual wattage used, there are higher IR losses (little more heat generated by inverter) but power from your batteries are just slightly higher then actual bulb wattage.

Associated point, many think that the batteries are delivering a relatively smooth DC to the inverter. If you have a 60 Hz sinewave inverter with a resistive load on the inverter, the actual current waveform from the batteries looks like a 120 Hz full-wave rectified DC waveform. The RMS battery current times the battery voltage equals the actual power from the batteries. Ignoring inverter efficiency, the peak DC battery current will be 1.41 times the RMS current. Factoring in inverter efficiency will make the battery peak current (and rms current) a bit higher.

icarus · July 2009

Re: Power Factor correction with CFL bulb question?

No offense RC,

But you are so far over my head that all I can see is your feet!

I do get it sort of,,,,

T

mike95490 · July 2009

Re: Power Factor correction with CFL bulb question?

RCinFLA wrote: »

Associated point, many think that the batteries are delivering a relatively smooth DC to the inverter. If you have a 60 Hz sinewave inverter with a resistive load on the inverter, the actual current waveform from the batteries looks like a 120 Hz full-wave rectified DC waveform. The RMS battery current times the battery voltage equals the actual power from the batteries. Ignoring inverter efficiency, the peak DC battery current will be 1.41 times the RMS current. Factoring in inverter efficiency will make the battery peak current (and rms current) a bit higher.

I wonder if that would be an application for a " 1 Farad Super Cap " to reduce the "120 Hz surge" on the batteries - are they sensitive to that?

westbranch · July 2009

Re: Power Factor correction with CFL bulb question?

HELP.

this Forester needs a Philly lawyer to help decipher all that good info from RC
gulp...;)
Eric

john p · July 2009

Re: Power Factor correction with CFL bulb question?

I have never heard of a problem opperating CFLs from an inverter.. I have 12 x18 ones in my house quite often all on at same time.. its no problem as all are run from solar generated power.. and the inverter is a 400w modified sine wave. it never gets hot .. If I measure the amp draw from the batteries to the inverter I find each 18w CFL actually draws 24 watts..so not much waste power..
So no supercapacitors or eye of eagle or tongue of toad need to be connected to the battery to improve performance.

westbranch · July 2009

Re: Power Factor correction with CFL bulb question?

After a good sleep on it, ....

In a different thread, CFLS and MSW used together were discussed, in that they work together reasonable well. (I had thought that you would need TSW for all electronic devices incl CFLs.)

I am now wondering if the cost of TSW over MSW is worth it for running CFLs?

Given the cost of CFLs now, they are dirt cheap, and if they fail a bit earlier than they would if on a TSW is this a great problem? I think not. If one had a CFL that would not work, trade it out...

just a thought.

Cheers
Eric

john p · July 2009

Re: Power Factor correction with CFL bulb question?

im my experience over the last 4 years the CFL bulbs last as long on modified sine wave as pure sine wave.. as long as you use good quality CFL like Philips not the cheap ones they dont last the feeble electronics in them fail.
I started out using a pure sine wave then tried a modified sine wave as lower self power consumption. 280 ma for MSW and 900 ma for PSW it is significant if you trying to conserve battery power.

its the small inverter on the left the other one is pure sine wave for on demand use. computer and other things

BB. · July 2009

Re: Power Factor correction with CFL bulb question?

80%+ of the devices out there will work OK on MSW... You may have some 120 Hz buzzing (sharp edges in MSW can cause things to vibrate). You may have less efficiency by a bit. And if you have audio/video gear connected, you may hear/see more hash.

Some of the downsides... Make sure you don't "ground the neutral" of any inverter and ground reference the battery without reading the instructions first... Most (all?) MSW inverters will end up with a short if the battery and "neutral" are grounded together.

Most (all?) TSW inverters are OK with neutral/battery grounding.

My two cents--when I have purchased inverters over the years, they have all been MSW and I have only had a few problems with something not working right (typically wall wart transformer overheating).

However, if I was off-grid an relying on the Inverter--I would go TSW--at least a small one (couple hundred watts) to run my smaller/expensive/critical items. If there was an AC fridge involved--I would bite the bullet and go TSW.

If I had to run something big on occasion (saw, pump, etc.)--I would be hard pressed to choose between a TSW vs MSW.

Read the inverter FAQ's for more detailed discussions:

All About Inverters
Choosing an inverter for water pumping

-Bill

GreenerPower · July 2009

Re: Power Factor correction with CFL bulb question?

The issue with low PF is waste of capacity needed at the source. If the source VA rating is lower than the peak of the VA needed i.e wattage divided by PF, 50VA for a 25W at 0.5 PF for example, the sine waveform could be distorted (either voltage or current) then this would introduce additional energy lost at the load.

This is usually not much of an issue for inverter systems since they have much higher surge capacity than their VA ratings. But would be a real issue for electricity power providers. Over a decade ago when switching power supplies in PCs are not PF corrected, this introduced stress on the power plants since they need to run to provide enough VA. If 1 MW load of PCs at 0.5 PF, this means the power plants need to provide 2 MVA capacity and this costs more fuel (instead of running something like 1.25 MVA for 1 MW load at 0.99 PF). PC manufacturers are now required to have switching power supplies meeting certain PF level.

I imaging similar requirements are going to apply for CFLs when their usage hits some level.

On a side note, all devices with bridge diodes and capacitor connecting to an AC outlet would have poor PF. CFLs and switching power supplies would fall into this category.

These devices would operate similarly with a constant 120VDC source (i.e. internal bridge diodes are not really needed) as with 120 VAC . This explains why MSW works better with CFLs since they have flat tops like DC source in each half wave. After the bridge diodes, the MWS just look like a DC source connected about 80% of the 120Hz cycle and disconnected the remainder 20%. The filter capacitor just smooths this out to be an average 120VDC source. True DC source is even better for CFLs. But you would need something like a low resistance inductor in series to limit the surge current to charge the capacitor upon connection. Otherwise, this high surge current could destroy the bridge diodes.

GP

Cariboocoot · July 2009

Re: Power Factor correction with CFL bulb question?

This may be a stupid question but ... will this power factor problem show up on a Kill-A-Watt meter?

Oh, btw Tony: "up" is 118 kilometers, according to the University of Calgary. They say that's the point when you're out of Earth's atmosphere and into outer-space. Once you're there, "up" and "down" are meaningless (which is not to say you are out of Earth's gravitational field which is actually infinite as it is for all bodies).

BB. · July 2009

Re: Power Factor correction with CFL bulb question?

Very early production kill-a-watt meters actually overheated with MSW inverters... They made a change and that problem has been fixed. (somebody here called "P3" and found what the cut-off was--but I annot find the link now). Any meter less than a few years old should be fine.

I have two different laptops (one old, one new)... and the kill-a-watt shows a PF of 0.6 on one and 1.0 on the other. So, it does appear to do a decent job of measuring power factor.

-Bill

RCinFLA · July 2009

Re: Power Factor correction with CFL bulb question?

If you are running modified sine wave (3 step square wave) do not bother with PF corrected bulbs.

Some cheap bulbs have problem with modified sinewave inverter particularly at cold temps due to the created high voltage DC is a bit lower then developed from a sinewave 120 vac source.

Modified sinewave inverters have peak voltage of about 155 volts versus 169 volts peak for a sinewave.

westbranch · July 2009

Re: Power Factor correction with CFL bulb question?

RC, is there a way of identifying the PF corrected CFLs?

Eric

icarus · July 2009

Re: Power Factor correction with CFL bulb question?

Eric,

Use a Kill-a-watt. Some bulbs might also list a VA as well as watts.

T

westbranch · July 2009

Re: Power Factor correction with CFL bulb question?

sorry, should have expanded my Question,

is there a way of identifying the PF corrected or non corrected CFLs before you buy them?
Is there something coded on the package that identifies it (PF) ?

Price?
Brand?

Eric

mike95490 · July 2009

Re: Power Factor correction with CFL bulb question?

westbranch wrote: »

Is there something coded on the package that identifies it (PF) ?

Price would likely be higher (more parts) and they would likely have a PF corrected logo, or something, to show you why they are asking for higher price.

john p · July 2009

Re: Power Factor correction with CFL bulb question?

Can someone tel me what all the anxiety and thoughts about possible problems with power correction on CFLs is??.. YES IT CAN BE A PROBLEM ON SOME ELECTRIC MOTORS .
Buy good branded CFLs screw them into to light socket and turn them on and thats it.. see wotking perfectly.. worry about something useful its not like you are going to be using and having 50 of them turned on at same time.
IF YOU HAVE GRID MAINS MODIFIED SINE WAVE PURE SINE WAVE THEY WILL WORK AND LAST LONG TIME ON ALL 3..:cool:

BB. · July 2009

Re: Power Factor correction with CFL bulb question?

Unfortunately, PF issues can be a relatively big issue for off-grid applications...

For example, the MorningStar SureSine TSW inverter is rated at 300 watts continuous at 25C... In reality, it is probably rated at 300 VA (Volt*Amp) or:

VA = Watts/PF ; or Watts = PF * V * A

Say you have 10x 13 watts CFLs with a PF of 0.5 (bad, but I have seen worse).

Somebody looking at the specifications--they see a 300 watt inverter - 130 watts = 170 watts for other loads... A laptop PC at 60 watts and a 25 watt Halogen desk lamp looks doable... But--the PC is an old one with PF = 0.6 -- we add everything up in watts and VA:

Watts = 10 x 13 Watts + 60 Watts + 25 watts = 215 Watts
VA = (10 x 13 Watts / 0.5) + (60 Watts / 0.6) + 25 watts = 385 VA

So--if you add everything up in watts--you are running at 72% of the inverter's rating.

In VA, you are running at 128% of the inverter's rating...

Battery/solar panel wise--you are still running 215 watts (+ losses). But your inverter is actually overloaded at this point (may shutdown in hot weather -- or may not last as long because of elevated operating temperatures). Also--your wire is carrying more current than you planned on -- so you may have more voltage drop and/or more wasted energy than you allowed for.

PF issues have hit large companies before (decades ago, a local credit card processor using a whole building full of, then new, PC's kept burning out wiring and popping fuses--even though the the circuits were well under the power ratings--bad PF was the cause).

-Bill

john p · July 2009

Re: Power Factor correction with CFL bulb question?

Unfortunately, PF issues can be a relatively big issue for off-grid applications... BUTwe only talking about running a few CFL in a house.. not using enough of them to light up a football stadium...
My house has 12 on almost all the time as I said running of small inverter it never gets hot enough for its internal fan to come on and I live where average day temp is 90 to 100 deg f.

Cariboocoot · July 2009

Re: Power Factor correction with CFL bulb question?

Bill has quite accurately pointed out that this is a cumulative problem. The more devices you have with a bad PF, the worse it gets. If your system has nothing with a low PF on it, you'll never have any trouble. But when you're trying to squeak maximum efficiency out of an off-grid system, it's important.

It's also important to utilities, as they notice more power being used than they get to bill for. Oh, did I imply it had something to do with profits? Sorry! Didn't mean to!

icarus · July 2009

Re: Power Factor correction with CFL bulb question?

Bill has my situation down perfectly.

(We don't routinely run 10 bulbs at any one time,, but we could)

TS 300 inverter,, 4-6 cfls with Pf ~.5,, add in 2 lap tops @ 58 watts PF .51,, the modem @ ~20 watts PF .5, router 15 watts PF .5 etc and all of a sudden my inverter is working much harder than I think it is. (as you all well know,, I'm not going to do the math since we will end up with gibberish!

So in reality,, it isn't a huge deal, but it certainly is something to consider in the design phase of any project. Nothing is free.

Tony

RCinFLA · July 2009

Re: Power Factor correction with CFL bulb question?

I would have to agree with John P, don't get too upset if you are only talking about three or four 20 watt CFL running at any given time. You will likely burn more oil in PC electric searching web for them, or worse, driving further to a store that carries them.

Couple of points from an earlier thread.

1) Most PC's still sold today still do not have PFC power supplies in them. The Energy Star 2.0 of last year does not have much teeth. Within the next year or two should see much more switching power supplies with PFC. Texas Instruments and ST Microelectronics, to mention two manufacturers, have marketed integrated switcher controller I.C.'s that contain PFC systems. Regulation does not address PF for devices less then 75 watts which is where most CFL fall.

I bought a Lenovo six months ago with Q6600 quad core CPU (90 nanometer process). It draws 200VA / 120 watts during boot for PF of 0.6. I just build a computer using Intel motherboard with E7400 dual-core CPU (65 nanometer process) where I specifically bought a PF corrected power supply that draws 60VA / 58 watts during same boot process for PF of 0.98 . Just examining Newegg's large selection PC power supplies I would be surprised if PF corrected units accounted for 20% of the offerings.

2) The main concern from electric utilities about power factor is increased transmission line losses due to higher IR loss. A PF of 0.5 does not mean twice the generator power. Yes, there is reduction in generator capacity due to winding current and core saturation limitations but other then primarily making up IR loss it does not burn that much more fuel. Typical rough average for electric utility grid distribution is about 10% loss in the step-up transformers, transmission line wires, and step-down transformers getting to end user of the electric. Your meter spins to true power (watts), not appearent power (VA). You can see square boxes with two insulators coming out the top, sometimes stacks several in parallel, on mid-level (100-250 kV) lines power poles that are oil filled capacitors used to correct PF by electric companies. Electric motors are primary culprit which create inductive PF.

BB. · July 2009

Re: Power Factor correction with CFL bulb question?

Capacitors can do a good job of PF correction for induction motors (can be brought back to near 1.0 with motor+capacitor)...

With digital power supplies (non-PFC)--a capacitor cannot offset bad PF because of the non-linear current profile of those supplies (basically a sharp peak of current, vs the the offset of current vs voltage of an inductive motor).

-Bill

JohnMassa · January 2013

Re: Power Factor correction with CFL bulb question?

Hello Tony,

Here is my take on the power factor issue:

The Power Company usually has several transformers between their plant and the user's location. These transformers work well on resistive loads and less efficiently on capacitive and inductive (reactive) loads. As a result the Power Company notices that reactive loads are inefficient and use more apparent power. So they charge you more according to how bad is the power factor of the devices at the user end. For example, a 25W CFL bulb have a power factor of about 0.5 so its apparent power (Apparent to the Power Company) is 25W/0.5 = 50 Watts. So they charge you for the 50 Watts. If you have your own generator on site (No transformers or long transmission lines) your 25W CFL bulb should only use 25 Watts.

John

BB. · January 2013

Re: Power Factor correction with CFL bulb question?

Welcome to the forum John!

As always, the answer is (usually) more complex about power factor...

Regarding the utility, today in California (for example) they do not charge for VA (or kVAH) on residential power bills. So, your 100 Watt Equivalent 25 watt CFL with 0.50 PF will cost you (me) at the 25 watt * Hours used rate.

If, this was a commercial customer, the utility may put a meter on that measures the average PF peak over 15 minutes in a one month (or one year) period. Then they take that PF reading and take 1/PF times the kWH used. This was common at oil refineries and other locations that use large motors. And, it made sense for large commercial power users to install capacitors (switched with motors) to bring the PF close to 0.95 or so (1.0 is "perfect", but can run into other issues).

So, a 100 Watt lamp replaced with a 0.50 PF 25 watt CFL would be (effectively) billed at the 50 watt * hours rate. (Commercial billing is actually two parts, one is the kWH used, the other is a "reservation charge"--the 15 minute peak kW power usage in a 1 year period--Roughly 1/2 the bill is kWH costs, and the other is reservation charges).

Regarding designing your own off grid power system... More or less, you need to design the AC side of the power system based on VA (VoltAmp) ratings (i.e, that 25 watt CFL looks like a 50 VA load). The wiring and the Inverter/Generator AC output ratings (for smaller/non-commercial systems, the Watt Output Rating is usually equal to the VoltAmp rating).

However, on the DC Power Side/Generator Fuel usage, there you would use the Watt/kWH ratings to plan the DC wiring and fuel storage/generator engine size (there are a few percent more losses because of bad PF, but they are usually swamped by the normal power usage numbers).

But, you still have to be careful with DC power side... Inverters are "constant" power devices... So if the DC voltage falls, the DC current rises to meet the AC energy requirements (power=voltage*current). For example a 12 volt 1,200 watt inverter would at first look, seem to require about 86 amps during the day (1,200 watts/14.0 volts=86 amps)... However, when you take losses and a low battery bank voltage (plus wiring losses, and NEC derating into account):

1,200 watts * 1/0.85 inverter efficiency * 1/10.5 inverter cuttoff * 1.25 NEC derating = 168 amp branch circuit/breaker rating

Or almost 2x what one would have first guessed.

-Bill

PS: I should add that motors (inductive) power factor issues can be "fixed" by capacitors. And capacitive power factor issues can be fixed by inductors.

However, most non-power factor corrected electronic power supplies and CFL's with electronic ballasts cannot be "fixed" with a capacitor.

Inductors and capacitors are "linear" devices--they just shift the AC current sine wave ahead or behind the AC votlage sine wave.

Most non-PFC devices (power supplies, non-motors) are not linear devices. Their current usage is "spikey" -- They take current only near the peaks of the AC voltage and none at the "shoulders" of the AC wave form. Capacitors cannot be added to the wiring to "correct" these non-linear effects.

You either have to accept the the poor power factor, or purchase power factor corrected lighting equipment/power supplies (they are out there, at least they have been in the past).

niel · January 2013

Re: Power Factor correction with CFL bulb question?

john,
it is not the transformers causing power factor or out of phase power as even an inverter or generator connected directly to a cfl will still show a va and power factor not representative of a pure resistive load. the reason is because the load is reactive and not the generating source or transmission lines so much. you are correct that their lines and transformers are reactive to a degree and the power companies usually compensate for that.

JohnMassa · January 2013

Re: Power Factor correction with CFL bulb question?

Bill,

I stand corrected, A local 120VAC generator feeding a 25W CFL bulb with 0.5 power factor will pull 25W/0.5 = 50 VA. On a 120VDC generator or 120V battery the draw will be 25 VA. The AC generators are a better choice as many appliances cannot use DC.

It seems that multiple transformers and long transmission lines should cause more losses due to poor power factor than does an on site generator that doesn't have losses due to multiple transformers and long transmission lines. The main losses must be primarily in the generators?

John

John

Cariboocoot · January 2013

Re: Power Factor correction with CFL bulb question?

John;

In fact you are right that transmission losses can be enormous. The losses can be 50% in fact. Power companies take this into account and write it off as operating overhead losses.

But Power Factor typically applies only to the difference between the VA supplied and the VA used by the end load. Not an issue for on-grid households as they are not charged for this difference. It can be an issue off grid as it needs to be considered when picking an inverter and/or generator (which although rated as Watts really supply power as Volt Amps; they're not 'good' with poor power factor loads).

Power Factor correction with CFL bulb question?

Comments

Categories