Tesla App Shows 50% Solar Production vs SolarEdge & Utility

[AD3T]

Hi. I had 16.2kW Tesla "XL" solar panel & 2 Powerwalls (Gateway 2), all installed by Tesla a few months back. I'm in Las Vegas. Finally got PTO from NVEnergy on the last day of September, which was the last day of the high time-of-use on-peak summer rates (hah).

All has gone pretty well, took several visits from the PW team, but the crews were really good & generally no complaints other than communication -- and even that improved during the course of my order process.

Anyways, I have both of my inverters connected via ethernet -- though I also have the Zigbee "Neo" gateway plugged in to ethernet & power & sitting on the indoor side of the wall with the inverters on the exterior side of the same wall. I requested SolarEdge monitoring and after about a week Tesla gave me access to it. All has been working great, both Tesla app and the SolarEdge monitoring, no real complaints [a few feature requests, but will ignore those].

Today, I'm looking at my Tesla app and I see approximately half the production that I'd expect -- indeed, my Tesla App is only reporting about 26kWh produced today, while yesterday was 50kWh & all prior days range between roughly 45kWh to 55kWh [slight variance based on weather, ie cloud cover & temperature]. I then logged into the SolarEdge monitoring portal, and it's reporting my solar production for today as being 50kWh -- precisely what I'd expect.

I unplugged the black "Neo" gateway & plugged it back in -- it's actually showing a "No Signal" light on the front, which I believe is new. Given that I have the 2 Powerwalls & the Gateway 2, my understanding is that the Neo device isn't even actually needed (at least here in NV), and that the box is transmitting data itself [presumably via wifi]?

I know that recently Tesla changed things to allow the inverter reporting/data to be passed through to SE, so that you can get both SE monitoring data & data in the Tesla app, and without having to hardwire ethernet at all (my understanding is that this already worked for Powerwall installs but not solar-only installs). Nonetheless, since hardwiring & getting access to SolarEdge monitoring, my data has been accurate in both the Tesla app & the SolarEdge portal -- until today, when suddenly my Tesla app is reporting half of what I assume is the actual production (I won't know for sure until I can check my utility's reporting, which lags behind by ~2 days).

Any idea why the Tesla App reporting would be still showing data, but only showing HALF of the solar production? The other numbers (total household usage, Powerwall input/output, less sure about grid-export though) logically seem about right based on historical numbers, though I don't really have a way to verify them for sure.

It's strange to me that the Tesla app would still have data reporting, but not all of it -- what could cause this, any ideas? =)

Ty!

 

[AD3T]

Just in case it matters or is helpful:

Tesla iPhone app version: 3.10.9
Powerwall version: 1.50.2​

I do use a third party app on my mac to display some basic solar/PW stats [the "power flow" UI basically] on the macOS menu bar, but it's just accessing the data from the Tesla API for visual display on my Macs -- I see the same data in both the iPhone app & the third party
menu-bar app / they match exactly.

 

[Dypkny845]

I think I'm noticing something similar. I'm thinking the gateway loses "connection" for some time and when it does it doesn't count the #'s accumulated during the lost connection.

 

[BrettS]

If your inverters are connected to ethernet then the zigbee gateway will always say no signal. The inverters won’t try to connect via zigbee if they have an ethernet connection.

However, that shouldn’t be the issue. When you have powerwalls the tesla app gets it’s solar production data from the TEG, which doesn’t need the zigbee gateway. It could be a configuration error that would cause half your production to be reported, but it seems odd that it would be working one day and then have a problem the next. You could try resetting the TEG, but my guess is that you’ll need to call tesla on this one.

 

[AD3T]

How would I reset the TEG? Something I can do via browser to its IP and/or hostname, or via flipping breakers?

Indeed, it seems strange that it worked fine previously -- for all data prior to today, my Tesla app is completely consistent with the SolarEdge reporting (+/- 1%-ish at least) -- which had me thinking it's a connection and/or software issue, perhaps one of the inverters stopped reporting, but not really sure.

Definitely would like to try resetting the TEG tomorrow. I figure that if it continues to be off by 50% by mid-day tomorrow then I'll give them a call, but kind of hoping it's something I can resolve locally; perhaps I should even test & re-terminate the ethernet cables, though seems a little unlikely, but have seen weirder things heh.

 

[gcsutton]

For what its worth i just noticed a firmware change to 1.50.2. Today getting half of yesterday and today was perfectly clear and yesterday had high broken clouds. Pics show the issue. The peaks and widths of the sine waves are the same.

 

[Redhill_qik]

The app can't seem to count correctly at times. My reported power generation yesterday was also about 60% of the prior day. Going into the grid view and scrolling down showed more energy going back to the grid than the solar view reported was produced.

 

[AD3T]

Interesting. I'm guessing that it's purely a software issue since it sounds like others are seeing similar. The bit "weirder"/more perplexing question is, where exactly is that energy going -- it's either going to my storage or to the grid, and I'm guessing it's the latter, but I can't really tell for sure for another day or two when my utility updates their [very detailed, 15-minute interval] data/reporting.

Somewhat hilariously, today I'm actually seeing the Tesla App reporting solar production as 2X higher (!!) vs the SolarEdge data for today -- essentially the exact opposite of the problem I saw yesterday, it appears that Tesla App is straight "doubling" the data today instead of "halving" it like it was yesterday. Makes me think it's software, but who knows.

Will keep an eye on it today and go from there -- if it's still wonky this afternoon then I'll go ahead & give Tesla a call.

Will update the thread this afternoon, too. This forum is a great resource BTW, have been a lurker for years now, glad to have finally joined the friendly fray!

Ty for the replies & advice, I appreciate it!

 

[getakey]

I've never paid attention to the amount in the app, just looked at the curves. After looking I see on 11/2 the app said 17kWh, but Solaredge has 25kWh. The other days in Nov are consistent between app and inverter.

 

[wjgjr]

I've never paid attention to the amount in the app, just looked at the curves. After looking I see on 11/2 the app said 17kWh, but Solaredge has 25kWh. The other days in Nov are consistent between app and inverter.

Check again after sunset tonight to see if the 11/2 number changed and/or note the 11/3 number and see if it is different tomorrow morning - that is the bug I have experienced.

In any case, I agree that these seem to be more bugs with the app - generally if you download the data using the link in the app, it seems like that has the correct data.

 

[AD3T]

Thank you -- I'll check again after sunset tonight.

Quick update, I just noticed that my Tesla app numbers for today are suddenly matching the SolarEdge data, apparently starting sometime in the last hour or so. In the "Performance" tab of the Tesla app, when I look at "Yesterday" it's still wrong/showing the halved numbers [vs SolarEdge], but the "Today" data has become accurate again.

FWIW, I don't actually check the Tesla app like a hawk as it probably sounds -- but I do check the macOS menu bar app very frequently, mostly because it gives me instant data on household usage & helps me to go tell my wife to turn off all the lights she's left on (half kidding).

Really though, it's just very helpful, and I'm not trying to promote it by any means (though I do think it's great, and am a software developer myself but I have no affiliation with the app or its dev, etc). I'm basically not much of a phone-app kind of guy, but I do like to keep an eye on production, usage, and battery status -- right now my electricity is quite cheap ($0.05 to $0.06 / kWh), but I'm essentially running "test trials" for next summer (June to September) when our time-of-use plan is extremely expensive for 6-hours every weekday ($0.43/kWh from 1PM to 7PM), basically "learning" the ebbs & flows of our usage vs solar generation/PW capacity & formulating my plan to no longer have $800++ bills in the Vegas summer heat. Plus, I'm a data nerd. =D

Thanks again -- will continue to update with findings / as & if things change.

 

[gcsutton]

I talked to CS today. Level 1 agreed it looks like a problem.

I also downloaded the data, and integrated the area under the curve, confirming its the kWh total calc that's the problem.

Also CS confirmed I got 1.50.2 yesterday. Might be coincidence, but thats the day the totals went bad. They are also bad today.

 

[ArizonaJon]

I am still on 1.50.1 version. But I typically see the SolarEdge report versus the Tesla having a difference of about 5%. With SolarEdge 5% higher than Tesla.

 

[BrettS]

I am still on 1.50.1 version. But I typically see the SolarEdge report versus the Tesla having a difference of about 5%. With SolarEdge 5% higher than Tesla.

It’s been that way for me from the beginning and I think that’s normal. If you have a system with powerwalls then your solar production will be measured different ways by the solaredge app and the tesla app. The inverters directly report their production to solaredge and the tesla app measures production with a CT on the AC output of the inverter. I suspect that it’s just a slight difference in the different meter calibrations, although I have seen someone suggest that Solaredge is monitoring DC production, which would be higher than AC production because there are some losses in the DC to AC conversion. I’m not totally convinced that solaredge is monitoring DC production though.

 

[AD3T]

Yeah, the slight difference (maybe 5% -- relatively small though) between SE/inverter reporting & the Tesla app's data has been that way for me the entire time, and I similarly attributed it to either DC-vs-AC or potentially even just slight discrepancy/inaccuracy in the way the CTs / Neurio devices measure vs the way the inverters measure.

I don't have enough data quite yet, but eventually I'll do some comparisons against my utility's reporting which, I would assume/hope, are probably the most reliable numbers [and even then, likely a non-zero amount of line loss occurs]. As long as the numbers I'm in the reporting [all sources] seeing are +/- 5-10% or so of each other so, then I'm happy.

Interestingly, my Tesla App is actually still showing the "halved" number -- but my macOS 3rd party menu-bar app is actually showing the full/accurate solar production number (ie it agrees with the SolarEdge portal). That leads me to conclude that the issue is certainly software related, and likely there's some kind of "mis-match" between what the Tesla API is sending vs what the client [iOS app] is expecting it to be sending, or similar along those lines (the Tesla API could even just be spitting out bad data, though seems unlikely since third party app began reporting accurately [and differently from the Tesla iOS app] in the last few hours).

Seems like it's probably related to the recent software/firmware update -- glad to hear someone called in, I'd expect that Tesla will figure it out and get it fixed en-masse pretty rapidly, but will continue to post updates if anything changes.

 

[getakey]

Check again after sunset tonight to see if the 11/2 number changed and/or note the 11/3 number and see if it is different tomorrow morning - that is the bug I have experienced.

In any case, I agree that these seem to be more bugs with the app - generally if you download the data using the link in the app, it seems like that has the correct data.

You aree right. Now 11/2 is at 25.6 kWh and 11/3 dropped tto 16.5 kWh

 

[wjgjr]

You aree right. Now 11/2 is at 25.6 kWh and 11/3 dropped tto 16.5 kWh

I have had this bug for months that data from exactly 3 days ago (11/3 for today) is constantly low in the morning, but seems to go up at times during the day, and becomes correct again by the evening. Then it repeats the next day for 3 days prior from that. No clue why, and it doesn't seem to be a bug everybody has - in fact, I'm not sure anybody else had confirmed seeing it.

It's not a huge thing since the underlying data and what I pull from the API is correct, but it is mildly annoying.

 

[ArizonaJon]

wjgjr, you said "It's not a huge thing since the underlying data and what I pull from the API is correct, but it is mildly annoying." Please help me understand were you are getting the "underlying data" and the "API" data. I have the data from the SolarEdge and from the Tesla android app. Is there another data source I should be aware of (other than the IOs app)?

 

[wjgjr]

wjgjr, you said "It's not a huge thing since the underlying data and what I pull from the API is correct, but it is mildly annoying." Please help me understand were you are getting the "underlying data" and the "API" data. I have the data from the SolarEdge and from the Tesla android app. Is there another data source I should be aware of (other than the IOs app)?

Underlying data does refer to the "Download My Data" link on the app - I am on android, but I assume IOS is the same. If I download that data for a date with "bad" data, the downloaded data is still correct.

As far as the API, there is both a local API and a cloud-based one. I use the local one, just to pull/store the current, raw data. There are much more advanced use cases, including being able to control aspects of the system behavior. (My understanding is with the newest software, this has become more difficult with the local API.)

 

[ArizonaJon]

wjgjr - Thanks for your response.