I have noticed than when our house is empty the load measure by the inverter is lower when the sun is shining 85W and higher 105W when the house is running from battery. The web portal shows these numbers. The Inverter app shows similar numbers.
Can anybody explain why this would be so. It seems like a current clamp measuring error but why would it depend on the power source?
nophead Is this anything to do with the power needed to run the inverter from the battery?
nophead the load measure by the inverter
The inverter doesn’t measure house load, it calculates it based on import - export + battery discharge - battery charge + solar generation
So when the sun is shining or the battery discharging you’ve got more power required to operate the inverter.
I think the battery discharge is measured on the DC side, so any losses on the DC to AC conversion will show up as extra house load
So its only an approximation
The Inverter app shows the power to run the inverter separately though. It seems like one of the sensors isn't accurate.
nophead The Inverter app shows the power to run the inverter separately though. It seems like one of the sensors isn't accurate.
that's unfortunately just a different approximation. I don't believe this is ever measured anywhere accurately
Most likely the grid meter is not accurate. In settings, try setting CT Correction to 1010 (or maybe 990) and see if that helps.
I think you have a single hybrid inverter? Plant setups and AC coupled are a bit more complicated. In the minute-by-minute stats, the 5th column is how much is coming out of the inverter, and the third column is how much is being exported (negative for import). House load is simply inverter minus grid. So if the grid reading is over estimated, the house load will seem less when exporting.
Yes a single Gen3 hybrid. Above 1000 increases the home load while running from solar so I will try to split the difference and see if it stays the same when the sun goes down.
Actually that setting appears to make no difference at all. Changing it from 800 to 1100 doesn't change the house load while exporting 3kW.
nophead That's odd. I'll do some testing tomorrow when the sun is shining.
Seems to work fine for me. I had the EPS load on backup for testing. With CT Correction default at 1000.
Then with CT Correction set to 1020.
The house load reading increases but no actual change in load.
That is a 37% increase but you only changed it by 2% or do I not understand it?
The house load is never measured, it is calculated by measuring the output of the inverter and subtracting the amount being exported (because the power has to go somewhere and it can only go to home or export).
When the sun is shining and the battery full, maybe 3000W is coming out the inverter, and 2900W is being sent out to the grid. Therefore the house is using 3000 - 2900 = 100W. But these measurements may not be accurate. Just a 1% inaccuracy could mean 3030W is really coming out the inverter and 2871W going to the grid, so then the house load would measure as 3030 - 2871 = 159W, Alternatively 2970W could be coming out the inverter and 2929W going to the grid so the house load would be 2970 - 2929 = 41W. So even with measurements 99% accurate, all we can say is the house load is somewhere between 41W and 159W. And I'm not even convinced GE measurements are even 99% accurate!
In the example, the CT correction adjusts the grid figure by about 2% (from 2665W to 2612W) so the missing 53W has to be added to the home figure (increasing 143W to 196W).
CT correction only adjusts the grid figure, so at night in eco mode (when grid is zero) it will have no effect.
So in answer to your original question, the night figure of 105W is the most correct. During the day measurement errors are giving you the false 85W figure.
Thanks for the explanation. I have it set to 1100 so it should reduce the grid figure by 10% and add that to the house load but it doesn't seem to affect it. For example it is exporting 314W at the moment but it showing a house load of 100W, when it should have added 31 to it. Even when it exports 3kW the house load goes down to the high 80's. It should go up by 300W.
Perhaps small values do have an effect but it doesn't seem to take effect for hours. I set it back to 1000 but it says my house load is 159W, which is too high. The GE app is showing 87W. So it seems like it is using a value I tried hours ago.
That's very odd, maybe there is a bug in my program. When I try it, the effect is immediate and it settles in a few cycles.
I wouldn't worry about it too much. There is nothing wrong, just measurement inaccuracies in the GE equipment. Your inverter is just a domestic grade product that works about well enough. A precision scientific instrument accurate to within a watt would probably have a million pound price tag!
Why does it take a few cycles. Is there some filtering on the displayed values?
I like an accurate load value because when we leave for 90 days I need to be sure everything that can be is switched off.
The accuracy of any app measurements (GE, BBCbasic or Inverter app) will come down to the accuracy of the GE equipment which, as has been said, is consumer grade specified for the purposes of solar energy conversion and battery storage.
I have found an alternative system from Emporia (on Amazon) that has its own current clamps and smart plugs that has a greater accuracy. It seems to work by taking multiple measurements during the 50 Hz cycle and integrating the resulting measurements to produce a much more accurate measure of the power draw. You can use it with a single clamp just on the house feed but it can measure every individual circuit in your distribution board (with a clamp on each circuit) as well as each appliance through its smart plugs. Below is an example of its output...
Thanks I might get one. I don't need absolute accuracy I just need it not to depend on my export or import rate.
Hey
I have seen something similar.
I have had a gateway and 13.5kwh AIO installed in November.
Until 2 days ago all Giv mobile app data has coincided near perfectly within 3-5% easily put down to losses in Octopus graphs and solaredge appdata.
What changed 2 days ago was the first time I setup the AIO to export power using timed discharge.
The higher I set my export rate on the aio. Load in the app increases whilst load on the web portal shows a lower more normal figure.
Since doing the export all my graphs in the app are throwing very weird numbers also. Again the giv web portal seems close to the other octopus / solar edge data.
I will contact support and provide all the evidence I can however interesting I'm not the only one seeing this.
Edit added some screenshots.
App showing 347. Portal at same time showing 178.
Graphs in app showing obscene numbers. Portal shows actual generation.
2nd edit.
I think I have figured out some the discrepancy.
The givenergy web portal seems to get it's data from the CT clamps and pulling the load value from the gateway load side. Which seems to be taking losses into consideration.
However the mobile app data isn't calculating house load off the c.t clamps / gateway data directly but instead just taking AIO output and removing the grid CT clamp value. Which results in a higher figure load in the givenergy app.
Thanks
Bob
Bob_Bobbings What changed 2 days ago was the first time I setup the AIO to export power using timed discharge
I've just checked on my GE app and it is showing the same thing, but I do not currently run timed discharge (the time slots are set up, but the enable DC discharge is left off) - not sure if this detail helps