Solar tracking lower than inverter report

[fluffy_burrito]

For my solar production, I’ve noticed that Sense consistently reports 1.5-2 kwh lower per day than what my inverter reports. Is this normal?

[kevin1]

I did some detailed analysis of my SolarEdge inverter vs. Sense here:

https://community.sense.com/t/does-anybody-have-any-cool-graphs-or-analytics-using-sense-data-please-share/3302/4

Bottom line is that Sense came in approximately 4% under my inverter for any given day or hour, and Sense was more accurate when looking at comparisons against my revenue grade utility net meter.

So your Sense is probably more accurate. If you have access to .csv’s for your hourly production from your inverter, you should chart a comparison.

[danderson]

mine is about 10% lower then what my inverter system reports. But sense more accurately matches up with my bill from the power co

[fluffy_burrito]

I did a quick daily plot of inverter vs Sense with the data I have (9 days), and see that Sense is about 4% under:

This is from a 8kw array using APsystems micro-inverters.

[steveranthony]

I have a Solaredge system also… Sense does come in a bit lower, bit I suspect it’s within tolerance… IE… today I produced on Solaredge 9.6kw… Sense comes in at 9.1kw… I think that’s around a 5% difference… acceptable…

[kevin1]

I decided to do one more round of analysis with data from after my big April reset… 65 days of data.

First try - ‘Energy In Interval (kWh)’ is from my SolarEdge inverter, ‘kWh’ is out of Sense…

A quick look at the residuals (differences) vs the linear model identifies the two troublemaker points, 1 and 6.

Point 1 was the first partial day after I reset my Sense history so of course the Sense number is going to come up short. Point 6 is for data on April 11. My April 11 with Sense looks like this:

I had to do a hardware reboot on my Sense that day - not sure why it went crazy.

Pull out those two data points and the curve looks like this:

The linear model looks like this:
Coefficients:
(Intercept) Energy In Interval (kWh)
Sense kWh = 0.07832 + 0.96987 * SolarEdge kWh

So the drop-off from my inverter to Sense appears be closer to 3% nowadays…

I should probably take a quick look at the other points that fall slightly off the line.

If I take a look at the residuals after removing those first two points I see a few more that fall quite a bit out of line. Indices 4, 11 and 58 are up next for examination.

Index 4 is April 10th - That was just the start of the Sense monitor issue that continued into the 11th. Again, that was fixed on the 11th with a reboot.

Index 11 is April 28th - I had a short Sense dropout that day that lowered the Sense kWh number.

Index 58 is June 14th - A string of very short Sense dropouts

[fluffy_burrito]

Interesting. It would seem that Sense is always a few percent points under what the inverters report.

[danderson]

Esterillos 7.000kW

i also have APS Inverters, and my monitor for APS is usually about ~10% higher than sense. But sense is in line with my power bill and meters

[fluffy_burrito]

Interesting site. Are you somehow automatically uploading your solar data into it? How is that done?

[danderson]

via python script and some api calls… the info is on this forum at PVOutput.org Push - #13 by danderson

[suddenlygod]

My sense also reports a bit lower than inverter. Sense reflects power company almost exactly.

[dwleckie]

I have a 14.95 kW system - 46 panels with 3 Delta inverters. My Sense reported production numbers are almost always within 0.096-1.02% lower than the inverter reported output - barring Sense occasional communication glitches.

[asheys1]

I too have a SolarEdge inverter (49panel 15kw inverter) and noticed the lower total output per day but I also noticed that Sense is directly in-line with what is reported by my utility power meter. I just assumed that the SolarEdge is showing the kw being produced BEFORE the loss involved by the inverter and what is ultimately output by the inverter and being picked up by Sense and the utility meter.

I could definitely be wrong as I haven’t done much in the way of research but it seems logical. It’s good to see that my findings are inline with what you are all reporting.

[kevin1]

There are probably a couple of loss sources between the inverter measurement and the grid-tie where Sense is doing the measurement:

• The inverter itself probably chews up some power.
• Then there are ohmic/resistive losses from the inverter to the grid tie.

[fluffy_burrito]

Something clicked for me just now. We’ve established that Sense is consistently reporting about 4% lower than what my inverters report. My inverters’ efficiency is about 96%. So what’s probably happening is that the inverters are reporting DC power generated, whereas Sense is reporting AC power generated. That would explain the ~4% discrepancy.