Weird Tesla App Anomoly

[wjgjr]

Anybody else notice this odd behavior in the Tesla app? I have been watching my monthly production numbers in the app, and I've noticed the last three mornings that my August production has dropped 2 or 3 kWh from the evening to the next morning. So, yesterday, I wrote down the daily numbers from the app. Then, last night I checked them, and one specific date, 8/9, had gone up, from 28.3 to 30.6. This morning, I checked, and 8/10 has gone down, from 34.2 to 30.9. So it seems like for some reason, the data from three days ago is dropping overnight and then going back up in the evening.

Since I am logging all of the data, I can confirm that the higher numbers are correct and all other days in August perfectly match between the data I logged and the app. So it is more an oddity than anything, but curious if others have seen it. I am going to try today to see if I notice when the 8/10 data goes back up and also whether tomorrow morning I find that the 8/11 data has gone down.

I am on the latest app version for android.

 

[BrettS]

I certainly haven’t noticed anything like that, but to be honest I don’t often go back and check previous day’s production like that. I have an iPhone, but I’ll make a note of the last few day’s production this evening and then check it again in the morning and see if I notice anything weird.

 

[wjgjr]

Just to add to this, on 8/2 I posted in a different thread about my July usage, which I got from the Tesla app, and I notice now that number also was slightly low (by 6 kWh), so I suspect the same behavior was happening then. My June usage is still unchanged from what I stated on 8/2 (which makes sense if it is consistently having an issue with the data from three days ago.)

And, I just looked again, and my 8/10 number (which I mentioned above should be 34.2 and was 30.9) is now 30.0. So it is actually changing over time (not just a single drop.)

 

[Troglodytes]

I have noticed some apparent manipulation/modification of the data as well. iPhone/iPad app user.

In the first month after PTO in early July I downloaded data from the app on one particular day, was jotting daily data into emails to friends (sharing the coolness factor of the system), and also looking at the graphs/tables data at various times.

At one point, I noticed that the data shown in the app didn't match the data that had been downloaded previously nor the emailed data values either.

I chalked the issue up to back-end changes by Tesla.

But.. if I compare daily production data for August so far between SolarEdge and Tesla I do see a 3.7% aggregate difference so far (Tesla app says 927kwh while SolarEdge says 964kwh). Maybe inverter and/or resistive loss.

However, doing the same check for July the Tesla app claims 20% more production for July than SolarEdge (Tesla app July 1666kwh, SolarEdge monitoring app 1330kwh).

Something wrong here between the reporting of the two apps for the July time period. Definitely some disagreement between the apps.

My system is owned. And, I have the black square Tesla Zigbee to Internet device unplugged; so they likely aren't pulling data from my inverters for CT data comparison. But, I wonder if Tesla benefits from using 'different' numbers for performance based contracts. Not saying this is the case, but just a hypothetical. I assume the language of those contracts is specifc to how and where the data is gathered from.

 

[wjgjr]

This potentially is a different (and perhaps bigger) issue than what I was seeing, which I believe is just a weird display bug in the app. (Side note, my 8/10 data now shows 32.5 kWh after a force kill of the app, so up from 30.0 but still below 34.2.) When I download the 5-minute interval data from the app, I consistently come up with the same number for that date (closer to 34.3 based on the instantaneous numbers the export shows.) I have not yet spent the time comparing the Tesla numbers to the inverter numbers.

 

[wjgjr]

Just to add to this - somewhere between about 5:00 and 6:30 yesterday, the 8/10 data went back to normal. But the 8/11 data, which previously (and correctly) reported 35.0, this morning shows 30.7. Guess I will have to report it to Tesla. All the underlying data seems good, so very strange that it consistently has a display bug specific to 3 days before today.