CFLs saving energy?

[John Rumm]

Thinking aloud here...

Regular readers may recall discussions of some (possibly short sighted) legislation introduced in Australia mandating that all new PSUs sold must be of a switched mode design (under the assumption this would reduce energy usage). As some annalists have argued, if enacted, this may actually increase overall energy use by lowering the efficiency of the distribution network.

So, does the same argument apply to CFLs?

Currently IIUC, there is no requirement for the majority of them (i.e. the lower powered ones) to be power factor corrected, even though many have truly horrible current consumption profiles. If/when there is a wholesale shift to these, such that they became a measurable percentage of the total grid load, will the extra supply waveform distortion created result in more distribution losses than the energy saved by the bulbs? (given that the losses will be a percentage of the total load, not just of the proportion of it attributable to lighting).

Discuss....

[T i m]

Can / do they not apply some form of power factor correction, like they do in companies running loads of machines etc?

Maybe there could be a market for a domestic PFCU, especially if it made the meter go round slower? ;-)

Cheers, T i m

[John Rumm]

Domestic customers are usually billed for Watt hours, rather than VA hours - so it would not save much if any money for domestic users[1]. Industrial users however are typically charged for the reactive current they use "use", so it it certainly worth it for some of them.

[1] The slight increase in harmonic noise generated within the properties wiring would result in a bit more real power being dissipated in all the mains input filters on the power intake of most appliances - but that ought to be fairly small at the household level.

[Harry Bloomfield]

T i m submitted this idea :

There are two methods of PFC, which is suitable depends upon the type of load. If it is normal for all the machines to all be running at the same time, then they can use PFC centrally. If various machines are turned on and off dependant on need, then each machine needs its own PFC.

In a domestic environment, where the number of CFL's in use will vary dependant on need - no standard load, then each one would need its own individual PFC.

[PeterC]

Most CFLs seem to be about 0.5 or a bit over. The Status 30W CFLs (from Morrisons) are about 0.98 or so. Unfortunately, Status' lower-wattage CFLs are also low PF and 30W is a bit high for most areas.

[NT]

The energy loss caused by pf =3D not 1 is trivial, and 1-2 orders of magnitude less than the power savings they produce.

As usual on electronic topics, the thread has been high in misinformation so far.

NT

[The Natural Philosopher]

yes.

Almost certainly not. There is sufficient residual resitance inductance in the lines and transformer feeding them to offset massive peak currents, and it slightly acts against inductive loadss like motors, which helps balance it all out.

(given that the losses will be a percentage of the total load,

Interesting theoretical problem..

>

[NT]

Its worth bearing in mind that the effects from this lowered power factor are quite different to the effects of lowered pf from inductive loads such as motors.

With CFLs, the rectifier reservoir arrangement results in the power drawn occurring around the voltage peaks. Since the lamp uses a switched mode psu, this results in less total i x t drawn, and less i^1 t, thus less copper losses.

With motors, the lowered pf means there is alternately power drawn and power fed back to the grid 100 times a second, resulting in higher max current draw and increased copper losses in the supply cables.

Its also worth bearing in mind that domestic supplies are split 3 phase supplies, so this higher peak current draw from CFLs occurs 6 times per cycle, not 2. Thus if the current draw spreads over at least

1/6th of the cycle, the generation sees a continuous electrical load from them. 3rd there is the fact that CFLs make up only a small percentage of total electrical load.

Between all 3 factors, the end result of CFLs having pf =3D not 1 is quite trivial.

NT

[John Rumm]

IIRC 30W is over the limit where they must have internal PFC...

[John Rumm]

So why do the power distribution networks go to quite some effort to control it?

Some numbers would be useful here...

What proportion of the national load is domestic?

What proportion of that is lighting?

If we know those we can make a stab at calculating the reduced load in a high CFL concentration alternative. (say 1/4 the real power, and 1/2 the current)

It may be that you get an overall reduction in energy consumption of a couple of percent. However ISTM that even a "trivial" loss of network efficiency, might mitigate some or all of that.

That seems a little rich coming from one of the key purveyors of hokey electronic ideas! ;-)

(although my news feed is a bit slow - so perhaps you are reading stuff I have not seen yet)

[PeterC]

Yes, AIUI it's >20W, but some ProLite crap I tried, nomibally 30W, were just as bad as most. Does a good PF (for a given wattage) reduce the heat from the lamp?

[Fredxx]

Any method of power factor correction is going to consume a small amount of power which will be dissipated as heat.

[The Natural Philosopher]

because it matters on big stuff.

about 30% IIRC.

probably less that 5%

probably has the figures you want.

No, the facst are that energy saved by CFLS is a very small faction of a small fraction. Its almost completely meaningless in the grand scheme of things. Thats where the ecobollox is bollox.

However the effect of the very very tiny leading power factor, compared with the massive industrial lagging power factor actually helps rather than hinders and is trivial compared even with the trivial amounts of energy saved. in short., its lots in the noise!

The fact is that CFLS wont save the plane., not in any way contribute towards it. Its pure spin.

However this particular argument against them doesn't hold water.

[The Natural Philosopher]

Not necessarily.

Capacitors are virtually 100% efficient at PF correction. At small power, and if you do teh rather complex mathemetics and simulate your diode junctions with the sort of logarithmic curve of voltage/current and do the time integral over a cycle to get some power loss figures out, you find that in fact being fed from a nice current limited inductive load with a better PF actually reduces power loss.

[Bolted]

Quick google for the latter, International Energy Agency reckons 19% of electricity use is for lighting, but that includes commercial and is a global not a UK figure.

[Fredxx]

I accept that capacitors are pretty efficient, but they will still run a couple or more degrees above ambient and the loss depends on the voltage waveform they are subject to. The more ragged, the higher the loss.

Oh dear, "fed from a nice current limited inductive load"?

[The Natural Philosopher]

well if you don't like that, do the engineering degree in electrical engineering, and I'll give it to you in the correct technical terminology.

AC + inductor=current limitation. Due to impedance of the inductor.

What about that dont you understand?

[Fredxx]

The bit that says "fed from a nice current limited inductive load".

A load is a sink of power, yet this load is seemingly able to feed some diodes. Perhaps you need a refresher course in technical terminology is that isn't what you mean. If that is what you mean, then I would suggest you redo your electrical engineering degree.

You might also have better luck than me in finding such a circuit. A typical CFL with an electronic ballast has a bridge rectifier feeding a switch mode power supply. Not many have mains side inductors. If they do, they'll be a small value to primarily reduce high frequency emissions, and will have little effect on power factor correction.

[John Rumm]

Agreed. How many thousands of CFLs added together constitute big though?

Sounds familiar...

So about 1.5% of the total load then...

Well indeed, but its something to keep people obsessing about while the big picture slides silently past.

I was thinking more in terms of harmonic losses than pure lead / lag PF issues - the CFLs effect may indeed be lost in the noise, however "noise" is what they will add presumably.

It was not an argument against as such, more just a question as to if one could quantify the effect.

[The Natural Philosopher]

Ok via a nice big inductor. Trying to make it simple.

Perhaps you need a refresher course in technical terminology is

leakage inductance on the substation transformer will be adequate.

quite a bit in the average house MCB as well. A

Granny, eggs ,suck.

see abiove

>