how to handle voltage fluctuation

Hi there, 

I have a general question about voltage fluctuation and measurement accuracy. I noticed that the voltage in my house come from 210 V to 245 V, based on the house consumption (high consumption -> low voltage, low consumption -> high voltage) and I don't know if this behavior is normal. BTW, the measurement change as well as the house consumption changes. I have done this trial: I measured a light when the voltage was 230 V and the measurement was correct (103 Watt. I tested it using a multimeter). As the house consumption increases, then the measured consumption of the light increases as well, up to 115-120 Watts. Obviously, also the reverse appears. 

I know that the error is very low, from 95 to 120 W, but I think that with high values the error increases.
The question is born because I have a pv system of 6 kWp and, sometimes, it measure 6500 - 6900 Watts.

Can you give me an explanation of that phenomena?
It seems that the firmware does not check the real Volts, but uses a standard measure.

PS: I use this firmware I used also this but the result is the same.

Therefore, I use the AC-AC adapter.


Robert Wall's picture

Re: how to handle voltage fluctuation

Voltage fluctuations are quite normal. They result from the changing load on the supply system.

Every component of your public electricity supply, the generator, the transformers and the cables all the way to your lamp, has a property called "impedance". There will be two parts to this, a physical property of the metal itself, the resistance, and for transformers particularly there is also inductance because the wire is coiled up on an iron core inside to form windings. Together, these impede the flow of current and in doing so a voltage appears across them. It is this voltage that is lost in the supply and causes the voltage at your lamp to drop when the greatest load is on the system. You've probably noticed that the voltage is high when most people are not using electricity, for example late at night.

Now your lamp (if it is an incandescent tungsten filament lamp) has resistance, if you measure it with a multimeter you will get confused because the resistance depends on the temperature of the filament (and that changes from say 20°C when the lamp is off to around 2500°C when it is on, and the resistance hot will be about 12 × the resistance cold, so it is a bad example. But if you measure a water heater (a kettle) the resistance varies little, because the temperature change is much smaller. The current drawn by your kettle is directly proportional to voltage - it is known as "Ohms Law", and the power is proportional to both current and voltage. Therefore if your voltage increases by 10%, the current will increase by 10% also, and the power will increase by 21% (1.1 × 1.1 = 1.21).

Finally, your firmware does use the voltage measurement to compute the real power. If you look at the sketch, you can see it happening in line 634:

634   long instP_CT1 = filtV_div4_CT1 * filtI_div4_CT1;

because "instP" is the instantaneous power, "filtV" is the voltage and "filtI" is the current.

All this is very basic electrical theory and you will find plenty of information on the Web and in text books.

hercules71035's picture

Re: how to handle voltage fluctuation

Hi Robert,
Thanks for the answer, it is very exhaustive.

So, if the (theoretical) power consumption of a light (or something else) is 100W it is not always equal. If the house voltage is high the overall power consumption is high as well. Am I correct?

Robert Wall's picture

Re: how to handle voltage fluctuation

That is (generally) correct.

There is one common exception: switched-mode power supplies such as you will find in your computer or TV. Those will probably be "constant power" because the power needed by the computer or TV does not change (or it changes, but not as a result of the mains voltage changing) and so the current will fall as the voltage rises. But these will be only a small part of your total load so overall you will see that the power consumption will be roughly proportional to voltage2.

hercules71035's picture

Re: how to handle voltage fluctuation

Thanks Robert,
you are great! :)

dBC's picture

Re: how to handle voltage fluctuation

That ohmic drop you see when your house is consuming 6kW will, all else being equal, turn into an ohmic rise of the same value when your house is exporting 6kW.  If you live in an area with widespread rooftop solar like I do, it's more common to see the voltage peaks in the middle of the day and the troughs at night.  In the chart below the vertical grid-lines represent midnight.

While everyone is at work and school there are a lot of empty houses dumping power out onto the grid, and that power might have to travel quite a ways before it finds a shopping centre, factory, school or office block to consume it.  Around here it's not uncommon to have 11kV feeders running backwards in the middle of the day.


The question is born because I have a pv system of 6 kWp and, sometimes, it measure 6500 - 6900 Watts.

That's probably not directly related to grid voltage.  I think a lot of inverters can go 10% over their rated output.  I have a 3kW inverter and 4kW worth of panels and when the conditions are right I see it output 3.3kW.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.