I am considering power supply similar to below schematic safely enclosed with my circuit in a box. Apparent power will be around 5-10W, but real power is well under 100mW. There will be several instances deployed, so reactive power adds up since each 10W costs approx $13/y
Somewhere I read residential customers are only billed for Real Power and only large industrial customers are billed for Reactive portion, can someone confirm ?
I believe a watt-hour meter measures true power
...but real power is well under 100mW.
I haven't analyzed the circuit, but I'm not sure I believe that. What's the zener voltage? What's C1? 220uF? 220pF? It's kinda funny you've got 1W resistors if you only expect to dissipate 100mW total.
Better question - why waste so much power to make an unregulated output of hundreds of volts?
230VAC thru full wave rectifier makes ~230 x sqrt 2 = 325V
You need better components than shown I think.
My mistake, it's actually 1W, not 10W, so not a big deal even if reactive portion is billed.
About components, I only looked up that schematic to demonstrate, I will use 1000V parts on 110V mains.
Is power really wasted? Reactive portion from any inductive load (e.g. ceiling fan, hair dryer, kitchen blender) will offset capacitive reactive power, will it not?
That 1000 uF capacitor will not see AC, anyway, so it will not be reactive power in the usual sense. It will have a sawtooth voltage and very "spiky" current waveform, depending on the load.
Or are you worried about the 225K capacitor?
A normal house does not have so much reactive load as to make a difference.
Factories with huge motors drawing lots of current with phase difference between current & voltage are what cause the problems.
My house has minimal motors - refridgerator, washer, dryer, water pump for furnace, and the occasional fan (window/ceiling fan, hair dryer, garage door opener, printer motor, computer fan). All small stuff.
Wallwarts these days all use switching regulators to make 5V, 9V, 12V, 20V for laptop chargers, and at pretty high currents. The big monster power bricks with huge transformers and full wave bridge recitifiers and hot-running regulators are fading away fast.
See:
http://www.mapawatt.com/2009/03/31/can-improving-power-factor-help-your-energy-bill
There is no way you will see a difference in a residential home when you try to adjust PF.
Thanks everyone. It's settled then, power utilities measure real power and residential households are usually slightly inductive (i.e. more inductive loads than capacitive) so adding 10-20W of capacitive power should not add anything to the bill 
Standard electricity meters are clever and are true power meters:
romanz:
Thanks everyone. It's settled then, power utilities measure real power ...
Not so fast. The utilities make certain assumptions regarding inductance. Big companies, commercial users, have "capacitor huts" and such to compensate, to bring the PF right.
It may be that the "smart meters" will end the assumptions they calculated for till now.
More heat than light? P-:
https://www.keysenergy.com/doc-download/PowerFactorFAQ.pdf
Find more. I googled
power factor and commercial electrical billing
Industrial user get monitored for the power factor they have. If it ever peaked over a certain value in a three month billing period they get charged at a higher rate per unit. This makes it important to them to have power factor correction to keep them on the lower tariff. This higher tariff is to encourage them to get the power factor down.
Domestic users are not monitored for power factor, so having a good power factor actually costs them more than not.
Some of those blogs linked confuse those two facts and think power factor correction saves money. It only does if you a penalised for poor power factor.
Most domestic users have quite a good power factor, for economy you don't need a power factor corrector but a power factor bugger er up.
Take the capacitor out of a floressent light and you run it vertically for free remove it from your vacume cleaner and that is free. However most domestic load is resistive so the actual impact on the bill is quite small.
The power compines do not want a high power factor as that increases dissipation in the generator. They monitor the power factor and if everyone started buggering up their power factor there are laws to stop you.
romanz:
It's settled then, power utilities measure real power and residential households are usually slightly inductive (i.e. more inductive loads than capacitive) so adding 10-20W of capacitive power should not add anything to the bill
That is pretty much the case. As Mike points out, industrial users are monitored for their power factor and penalised if it is undesirable. The "old-fashioned" power meters (no longer fitted to any installation as new or replacement) did measure "true" power, "smart" meters are now able (probably all of them) to monitor power factor in addition.
Your load here is capacitive, and not very "spiky" at all. KeithRB has in mind instead, switchmode supplies in which the current is drawn only near the peak of the AC waveform; their power factor is indeed poor but only slightly capacitive; though there are ways of correcting for this in very high power applications using a switching power converter on the input.
I made a post graduate course on harmonics (in english is Harmonics or armonics?)
The (controlled) rectifier circuits really create harmonics, making them uncasificable (from certain point of view they behave as capacitive but some of the aspects made them different). So do electronic regulators, being the (domestic) most important the light triac regulators.
For a single home it is not important at all (and the counter is not so intelligent to have this into account). The problem arises in big cities, where a lot of people have a big amount of rectifiers consuming (active) power in the same part of the AC cycle.
The apparent power obliges the supplier to have bigger lines, so they penalize low cos fi in a fixed basis; I don't have notice of an actual apparent power bill (this power returns to the system).
Regards.
(should somebody be interested on, I can tell you about an hillarant spanish XX siecle story on cos fi 
:D)
Hi, I have this type of smart meter.
https://www.powercor.com.au/media/1424/lg-meter-ref-card-rev-feb-2010-web.pdf
There are two LED on the front , one is "power usage" light, the other is not identified in the document but it has printed on the case next to it "VA"
It occasionally flashes, like once in 15mins I'd say when the house is really sucking juice.
It could be measuring the mysterious VA that everybody wants not to be measured.
Tom.....
Paul__B:
Your load here is capacitive, and not very "spiky" at all. KeithRB has in mind instead, switchmode supplies in which the current is drawn only near the peak of the AC waveform; their power factor is indeed poor but only slightly capacitive; though there are ways of correcting for this in very high power applications using a switching power converter on the input.
No, I am thinking of the circuit he has here. Assuming a DC load on the output, the only time the 1000 uF capacitor is fed current is when the AC Voltage through the rectifier exceeds the DC voltage (+ 2 diode drops!). This causes the AC line to spike the current to recharge the capacitor during the short time of the peak.
No, I am thinking of the circuit he has here. Assuming a DC load on the output, the only time the 1000 uF capacitor is fed current is when the AC Voltage through the rectifier exceeds the DC voltage (+ 2 diode drops!). This causes the AC line to spike the current to recharge the capacitor during the short time of the peak.
That's what I meant: this is neither capacitive (or "not exactly") nor inductive (nor resistive).
A lot of modern PC power supplies have really good PF. They don't use an initial smoothing capacitor.
In any case... what you'll see for current draw is the peaks minus the zener voltage and bridge drop. So it won't look like KeithRB's drawing.
I can't run LTSpice here, or I could show you. Anyone with LTSpice should be able to illustrate this.
We are talking about the current draw from the AC line, which is what affects power factor, this is indeed what that current will look like. Current will only be drawn when the AC voltage exceeds the DC voltage, which only happens during the peaks.
