CFL's and power factor

jeffkruse · June 2010

I did a search but still don’t think I understood the answer.

I have a CFL. On the Killowatt meter it shows 25VA, 13W, PF.5

How many watts is it drawing from my batteries?

Based on my search here I think it draws 13W from the battery but the inverter has to work harder, is that correct?

icarus · June 2010

Re: CFL's and power factor

There was an extensive thread on this very subject, I believe it was in the off grid section of the forum. Try a search at the top of the page. If not, I'm sure Bill with his magic fingers can (and will) point you in the proper direction,
(Edit) http://forum.solar-electric.com/showthread.php?t=4820

http://forum.solar-electric.com/showthread.php?t=5461

Post # 1 asks you exact question, post # 2 answers it!

T

A google search using your question plus "wind-sun forums" yields a much better result than the search function on the site.

T

jeffkruse · June 2010

Re: CFL's and power factor

Thanks. I read the posts but still don't understand.

BB. · June 2010

Re: CFL's and power factor

OK--it comes down to the difference between what a "Watt" is vs what a VA (Volt-Amp) is.

The fundamental equation for AC power is:

Power = Volts * Amps * Cosine of the phase angle between Volts and Current

Imagine you are pulling on a rope for a car. You can either stand right in front and pull, or stand off to the side (say 60 degrees of center).

If you want to put a 100 lb force pulling the car forward, obviously standing in from the of the car is the easiest way to pull. 100% of your pull on the rope will move the car forward.

Now, lets stand off to the side of the car (60 degrees to the left or right). You will have to pull harder on the rope because you are not only pulling the car forward, you are pulling it to the side too--but the wheels only let you move the car forward. The equation for how hard you have to pull on the rope to move the car forward is:

force-forward = force * cosine of the off angle
Force-foward = 100 lbs * cosine of 60 degrees = 100lb*0.5=50 lbs

To get 100 lbs pulling forward on the car, a little algebra:

force = force-forward / cosine X
force = 100 lbs / cosine 60 deg = 100 lbs / 0.5 = 200 lbs

So, you would need to pull on that rope with 200 lbs to get 100 lbs of force to move the car forward. Notice that the amount of work (force*distance) to move the car forward is the same in both cases, but pulling off-angle--the rope has to be 2x stronger to get the car to move the same amount.

So, here we have the mathematical definition of Power Factor for an AC circuit:

Power Factor = Cosine X

Now, the "phase angle" between voltage and current is complex (vectors--have both value and direction). And this stuff is frequency (and time) dependent too. Capacitors, inductors, motors, etc. all move the current forward (lead; capacitive) or behind (lag; inductive) the voltage wave form. When the phase angle difference between Voltage and Current is non-zero, then the Cosine of X or PF is less than 1.0). Very much liek the pulling on the rope example.

Now--to make it even more confusing... We have non-linear circuits too.

As an example, imagine that I stood in front of the car and pulled on the rope 200 lbs for 1 second and zero lbs for 1 second. I no longer use the Cosine X equation, but I can still have a "power factor" = 0.5 over time if I measure the jerky pull over time and average it out.

For CFL's, they are, typically, "non-linear" power users. For them, they draw zero current except for when the sine wave (voltage) is near its peak--then they take a "big shot" of current (usually a diode charging a DC capacitor to provide power for the CFL electronic ballast).

And yes, they can make electronics, motors, CFLs with PF=1.0 -- it just costs a bit more money. Some newer regulations are forcing larger appliances to have >0.95 power factor--and for large companies (like oil refineries), the utilities actually charge for "bad" Power Factor.

Now, how does that affect you...

Basically, your inverter is supplying "Watts" from the battery bank--so your battery is supplying ~13 watts.

However, the AC wiring, fuses/breakers, inverter have to be designed to provide the higher current (VA=25 Volt*Amps)... Sort of like the rope example--The rope has to be 2x as strong when pulling off-center of the car.

Power = V*A*PF
PF=Power/(V*A)= 13 watts / 25 VA = 0.52 power factor

Besides the wiring and inverter having to be larger because of "poor power factor"--There is a bit more power used/wasted... Because the current is higher, there is more heating of the wiring (and inverter transistors, etc.).

Power = I^2 * R

Resistive heating power is equal to current squared * the wire resistance.

Notice that if the "current is doubled" because of power power factor, the heating effect actually goes up by 2 squared (2^2 = 4) or 4x the wasted power... Another reason utilities and generators do not like "poor power factor"... More wasted energy.

For your home use, you are charged in Watts (kWatt*Hours). For large industries, they may pay in KVA (K-VAR -- Kilo Volt Amp Reactive). So they do pay directly for power power factor and will do things to make it better.

-Bill

AntronX · June 2010

Re: CFL's and power factor

Compact Fluorescent lights draw current at the peaks of voltage waveform. When measuring Amps, the peaks are included and not averaged. So, thats why Volts * Amps = 25 (VoltAmps). If current draw is averaged, then Volts * Amps(avg.) = 13 (Watts). The light bulb is still consuming 13 Watts, but it is asking for more current up front, and then takes a break untill AC voltage waveform changes polarity, then asks for another shot again. This behavior is best illustated in the image I attached. It is a picture of current (top wave) and voltage (bottom wave) waveforms of CFL viewed on an oscilloscope. Notice these peaks at the top waveform? These are the "shots" as I called them. In perfect scenario, both top and bottom waveforms would look identical and in step together. That would give you perfect power factor of 1. I copied this picture from here: http://cdonner.com/power-factor-and-compact-fluorescent-bulbs-cfl.htm

tallgirl · June 2010

Re: CFL's and power factor

In practice, the lousy power factor of CFLs may not affect the total "watts consumed" because the lopping off of part of the peak may cause them to be "on" longer.

But they are horrible devices and if they weren't so efficient, they'd be a mistake in off-grid systems. LED lighting =should= be better.

As an aside, I have an entire house full of CFLs -- standard three bedroom on-grid residence type of place -- and the load with all of them turned on is still =way= below what it would have been with incandescents. But the power factor is just miserable ...

BB. · June 2010

Re: CFL's and power factor

LED, CFL's, Florescent Lighting, etc... Power Factor depends on the ballast.

DC Powered LED's still need some sort of ballast (passive or active)--so there are still similar issues with LED lighting too.

You have have "cheap" ballasts which have bad power factor (non-linear wave forms) or somewhat more expensive power factor corrected ballasts.

The thing about electronic devices (and non-linear lighting) is that, unlike a typical AC motor, you cannot "fix non-linear" bad power factor with a capacitor.

With motors, for example, which are inductive and have "lagging" power factor, you can just put a (motor run) capacitor of the right size in parallel with the motor and the two add up to a PF near 1.0 very nicely. It will not save a home with utility power any money to speak of--so it is usually not worth adding PFC Capacitors--But it is nice when added in at the factory (allows appliances like A/C and Heat Pumps) to use smaller AC Circuits (less overall current).

-Bill

jeffkruse · June 2010

Re: CFL's and power factor

Thanks everyone.

The CFL might be drawing 25Watts for an instant but then the next instant it’s not drawing anything so the average is 13Watts.
Probably not a technically correct explanation because of the phase angles of voltage and current but I think I get it.

blackswan555 · June 2010

Re: CFL's and power factor

Inverters are in VA, Use VA for your calculations, I did a similar test some time ago, My 20 w CFL drew 37 VA, Not too much of an issue till you start adding them up & the hours in use,
A 40w incandes gave a better light for 3 more watts and far less initial cost,

Have a good one
Tim

jeffkruse · June 2010

Re: CFL's and power factor

blackswan555 wrote: »

Inverters are in VA, Use VA for your calculations, I did a similar test some time ago, My 20 w CFL drew 37 VA, Not too much of an issue till you start adding them up & the hours in use,
A 40w incandes gave a better light for 3 more watts and far less initial cost,

Have a good one
Tim

I have plenty of inverter capacity. I am trying to reduce the draw on my battery bank. So, as I understand it my “13Watt” CFL will draw 13Watts from the battery and not the 60Watts an incandescent would.

I’ll have as many as 6 CFL’s on at a time 13W*6 = 78Watts or 150VA. My inverter is a 4500Watt inverter so I have no concerns there. I just wanted to know if I was drawing (78W * 5hours = 390WHours) or (150W*5=750WHours) from my batteries. Either figure is still better than 60W*6*5hours=1800WHours for incandescent lights!

Now I just wish the CFL's lasted as long as they claim :roll:

BB. · June 2010

Re: CFL's and power factor

Regarding lamp life--If you have a MSW inverter, it may be causing your reduced life (if the electronic ballast is failing rather than the bulb loosing intensity).

If you can find a PFC corrected CFL -- or possibly replace with PFC standard tube fixtures may get longer life on the inverter.

Depending on how short the life is--wiring up a second, much smaller, TSW inverter just for the lighting may make economic sense (and recharging cell phones, etc.).

-Bill

mike95490 · June 2010

Re: CFL's and power factor

CFL bulbs need to keep the electronics in the base, cool. A base up bulb will fail very soon. Same bulb, base down , with good air circulation will last full life

BB. · June 2010

Re: CFL's and power factor

And 100% of my screw in bulbs are either base up and/or a "sealed" fixture (excluding a couple outdoor flood lights).

The only ones are not cannot use CFLs (oven, fridge, garage door opener).

I, for one, welcome our coming new CFL overlord and master... :roll:

-Bill

I guess I should add that I am pretty happy with my generic CFLs (other than color)... They last much longer than filament bulbs, much cooler in the home on a hot day, and usually last longer than those skinny tube under cabinet florescent lights. :roll:

a0128958 · June 2010

Re: CFL's and power factor

mike90045 wrote: »

CFL bulbs need to keep the electronics in the base, cool. A base up bulb will fail very soon. Same bulb, base down , with good air circulation will last full life

I don't believe this to be true anymore, as long as you purchase amalgum technology CFL bulbs (which may or may not be the least expensive or may or may not be commonly available in bix box h/w stores.)

I have 60+ CFL lamps in my home, with almost all 'base up.' None have failed so far since 4Q07. Attached is a slide from a recent speech I did, describing the bulbs I'm using.

Best regards,

Bill

blackswan555 · June 2010

Re: CFL's and power factor

So, as I understand it my “13Watt” CFL will draw 13Watts from the battery

Nope it will draw 25 VA from your inverter, translating , without losses , if your battery was at 25 volts to be 1 amp, Inverters work in VA not watts,

A good wiki on PF http://en.wikipedia.org/wiki/Power_factor

Have a good one
Tim

AntronX · June 2010

Re: CFL's and power factor

blackswan555 wrote: »

Nope it will draw 25 VA from your inverter, translating , without losses , if your battery was at 25 volts to be 1 amp, Inverters work in VA not watts.

Sorry but you are wrong. Without inverter losses, 13W CFL will draw 13W from battery. I connected 26W (50VA) CFL to my inverter and power draw from battery went up by 32W. In this case, inverter loss + measuring error = 6W.

russ · June 2010

Re: CFL's and power factor

The document referred to about PF is a generic document - nothing more.

I have CFL's in all positions - base up, base down and horizontal - have yet to see any difference the past three years.

There are a fairly wide range of colors available now - At first I selected the warm white range but have now tended toward a whiter color. Color rendering is not quite as good but I like the light better.

blackswan555 · June 2010

Re: CFL's and power factor

Ok,,,,,,,,,,, one last attempt,
Inverters and batteries do not work in watts they work in VA & AH , Unless you ONLY have resistive loads, You may as-well be doing your "watt" calculations in Mongolian love beads, I think the exchange rate is about as variable as VA > watts

Your Inverter "see`s" VA, You home electric meter "see`s" watts, (and possibly your test meter)

Have a good one
Tim

RCinFLA · June 2010

Re: CFL's and power factor

blackswan555 wrote: »

Ok,,,,,,,,,,, one last attempt,
Inverters and batteries do not work in watts they work in VA & AH , Unless you ONLY have resistive loads, You may as-well be doing your "watt" calculations in Mongolian love beads, I think the exchange rate is about as variable as VA > watts Your Inverter "see`s" VA, You home electric meter "see`s" watts, (and possibly your test meter)

Have a good one
Tim

Inverters do not work on VA, they are rated for maximum operation at VA.

It is their losses that depend on current draw during load. It is true there will be slightly higher inverter losses but at the power we are talking about for CFL it is relatively minor in comparison to inverter overhead, unless it is a very low power inverter.

Other tid bits,

Fluorescent bulbs put out more light when excited by 25-40 KHz versus old 60 Hz inductive ballast. A 40 watt T12 bulb running at 60 Hz and a 32 watt T8 running at 25 Khz (electronic ballast) put out about the same about of lumens. This is delivered bulb wattage and does not include the extra heating losses of an old 60Hz inductive ballast. It is very easy to convert these type of light fixtures to T8 bulbs (same base pins as T12) by replacing old 60 Hz inductive ballast with an electronic ballast. Most ballasts for these larger fixture bulbs are PF corrrected.

All CFL run the bulbs at 25Khz+ frequency. The weak link in CFL is the ballast design. Cheap ballast do not account for bulb temp causing higher drive when bulbs are hot. Most failure modes are induced from process of starting up the bulb. This is cumulative degradation. Rooms like a bathroom that light gets turned off and on a lot are the toughest on CFL.

Some do not do well at cold temps. Philips' bulbs are about the best for cooler temp loacations.

Fluorcent bulbs generate light with UV emitting plasma that in turn excites the bulb's phosphor coating. Different mixes of phosphor coating can eccentuate different color range. The further from blue end of the spectrum (more soft, more red, lower color temp) the less efficient the lumens output. For the same wattage, a soft white bulb will put out less lumens then a cool white bulbs. However, if you are trying to tell the difference between your navy blue socks and dark brown socks you might want the softer color, lower color temp, bulbs.

tallgirl · June 2010

Re: CFL's and power factor

jeffkruse wrote: »

Now I just wish the CFL's lasted as long as they claim :roll:

They actually =do= last as long as they claim, but that's averaged over more than a handful of them and for longer than a year or two. I've been testing CFLs for 3 1/2 years now, and I get an average life expectancy in excess of 7 years, and much closer to 10 years.

For raw numbers, I have 52 CFLs, of which 16 have been replaced in that 3 1/2 years. The problem is that they are "electronic devices" and "electronic devices" as a class tend to fail early, or late, but not so much in the middle. Unless you do a long term test, you'll see the early failures, but not the years and years of operation.

Dave Angelini · June 2010

Re: CFL's and power factor

I have several that have lasted over 15 years. If you have some extra pocket change buy anything made by Stecca. Even their cool white looks good! The warmer color soft whites look better to me than a halogen or incandecent.

The power factor really does not matter as there are few choices for lower power consumption at this point. Use CFL's, ALWAYS OFFGRID !

Cariboocoot · June 2010

Re: CFL's and power factor

The only CFL's I've had fail "prematurely" have been the encapsulated ones: those enclosed in a plastic case to make them the same shape as an incandescent bulb (necessary on a few light fixtures). I suspect this is entirely due to poor heat dissipation. I've even got a "loop" designed one here from way back that still lights (albeit somewhat reluctantly).

I'd expect poor performance from the el cheapo, no-name, made-in-you-know-where 'bargain' bulbs.

System2 · June 2010

Re: CFL's and power factor

Wow, whats all this Power factor malarky?

I use 12V DC lamps direct with my 12v sytsem and I always get years of use, never had premature failure. I guess this only applies to AC CFL's?

john p · June 2010

Re: CFL's and power factor

I've been testing CFLs for 3 1/2 years now, and I get an average life expectancy in excess of 7 years, and much closer to 10 years.Dont really understand this.. ????

For most people that have about 10x14 CFLs in a house the power factor is nothing to worry about on or off grid..I have been running 10x18w CFLs off a 400w MSW for 3 yrs no problems at all,,just buy good quality CFLs like Philips..

Cariboocoot · June 2010

Re: CFL's and power factor

solarboy wrote: »

Wow, whats all this Power factor malarky? I use 12V DC lamps direct with my 12v sytsem and I always get years of use, never had premature failure. I guess this only applies to AC CFL's?

Yes, PF applies only to AC inductive loads.
DC is great, but it can't do everything.

BB. · June 2010

Re: CFL's and power factor

With how Power Factor has been twisted from a linear function (cosine of the phase angle) to representing non-linear current flow--Depending on the actual DC CFL input current (steady state DC or pulse train)--You can run into "PF" type issues with DC too.

Because wire heating is P=I^2 * R -- a 50% current duty cycle will double the wire heating vs a steady state load.

-Bill

System2 · January 2011

Re: CFL's and power factor

Just want to tell of my experience in the area. I was a utility engineer and have been interested in energy conservation for some time. I installed a permanent 16 power channel logger in my home over 10 years ago and have been monitoring appliances, temperatures, etc. At this time I have no solar equipment installed but will considering it in the future.

In the last few years I've installed capacitors on various appliances and lights for the purpose of power factor correction some of which are now permanent, just to show it could be done and prove manufacturers could do a better job. I've improved the power factor on my refrigerator, dishwasher, washing machine, and some cfl's and circline lamp fixtures.

The first thing that I did was to build what I call a "capacitor box" that has plugs built in and a variable capacitance that could be installed between an appliance and the outlet w/o modifying the appliance to determine the capacitance required. Capacitors worked well on appliances with motors such as the dishwasher, refrigerator and washer (power factor like a cfl).

Capacitance also worked well on magnetic ballasted fluorescent fixtures used in table lamps to replace incadescents (circline lamp/ballasts w a 22w lamp that fits in a table lamp socket and a GE or Phillips Biax ballast for a 13w replaceable stick fluorescent) but did NOT work for electronic cfl's. Capacitance added to an electronic ballast or a TV set or a non high PF computer lowered the PF instead of increasing it. These were "Lights of America" ballasts with a Fc8T9 22w lamp that was sold at Wal Mart made in the US. Lamp life has been excellent on these.

What was successful was a 5uf capacitor measuring 3/8" x 1 3/8" x 1" which had a metal lug that I could attach to the top of the existing ballast with a screw. The leads from the capacitor were just soldered into the incoming wiring in parallel. What I ended up with was a change from a PF of .556 to one of about .0.98. Similar results with the 13w biax or PL lamp (phillips lighting catalog) except less capacitance is needed.

All other fluorescent fixtures use 2 and 4 foot T8 lamps w electronic ballasts. They are about 100 lumens/watt (better than the cfls by far) and have a PF of approx. 1.

Lighting manufacturers told me they didn't make the cfl's with a high power factor because they're too small. However my capacitor as noted above was small enough that it could have been done.

I've permanently corrected the refrigerator from 92% to 98% power factor by adding a 7.5uf capacitor. The heat pump was already about 90% and I've done nothing so far on that.

Just wondering if anyone else has tried something similar. I wouldn't recommend doing it unless you can measure power factor.

Thanks

niel · January 2011

Re: CFL's and power factor

juliotamu01,
welcome to the forum.

i don't think the issue here is whether or not a cap can correct a bad power factor or not, but whether it would be worthwhile to do it. utilities do not charge us residential users for a bad power factor and the only other time it may present itself is in the case of using a sinewave inverter and if the small gains of pf correcting would be worthwhile in such a case.

i do have an off the wall question if any can answer it.
will the gt inverters see a pf in connecting to the grid?
as i said utilities don't charge residential users so pf correction shouldn't be necessary, but was posed out of curiosity.

BB. · January 2011

Re: CFL's and power factor

Julio,

We had a big discussion of power factor and small generators here:

http://forum.solar-electric.com/showthread.php?t=10123

Only inductive loads can usually be corrected with a capacitor. And the capacitor should be switched on and off with the load.

Here is a thread where stevek measured the dc power used by an inverter with and without on a 120 vac induction motor. End result, a tiny bit of battery power reduction.

Regarding gt inverters, my utility meter can measure power factor. My gt system had a 1.0pf output.

From what I had read, utility companies actually have an issue with gt inverters and their 1.0pf output. Utility loads are typically much less that 1.0pf and gt inverters do not support the inductive part of the loads. So the utility needs to supply the out of phase current which they do not get paid for.

I would think that an inverter could be programmed to simply a 0.7pf output,but that makes the inverter more expensive because it would Supply more current for the same amount of solar power (more copper and bigger electronics to handle the higher currents).

-Bill

System2 · January 2011

Re: CFL's and power factor

Yes you're correct that utilities bill residential customers based on kWh rather than kVa however they are indirectly billed (by incrased $/kwh) for the increased I squared R losses due to increased transmission losses. I only did it to prove it could have been done and should have been done by the manufacturers of cfl's. Had the government mandated it, they would have manufactured high power factor cfl's.

Cfl's are often used in commercial and industrial applications as a replacement for incandescent lamps where the savings is lessened by poor power factor which they most likely have to pay extra for. In Houston, TX and other some other municipalities, the utility bills for kVa, not kW. In this case they likely would need to add power factor correction equipment at the service entrance that varies the capacitance. Most other utilities only penalize if the power factor is below 85%.

Those who are offline generating their own power by solar or other means that are not capable of correcting power factor at the source would benefit from high power factor appliances and cfls.

CFL's and power factor

Comments

Categories