I have been using Solcast from HACS. An update was released - v4.0.23 . I"m unable to upgrade from v4.0.22 . When I click update, HAOS just sits around and seemingly does nothing. It seems to fail due to a 404 on this URL, according to the log :
That’s sad. I certainly wasn’t one of them. Never contacted the owner.
I kept getting the upgrade notification, and it kept failing. Finally, I reviewed the logs and found out what was going on. I still think Home Assistant should have provided some kind of UI feedback that it failed.
You can delete it all though the gui. But not from the notification. That would be deleting the notification. It would also not be good UX to offer to delete any integration that has some kind of problem.
You can delete the integration on the integrations page. Not the right integration because I deleted mine, but same idea:
This is a crying shame as it worked well for me! It was just a case of making sure you stuck within your daily API limit and you called at a random time (as requested by Solcast). If I don’t touch it and skip the update will it just keep working??
Hope we can persuade the developer to bring it back.
It does not look like Forecast solar recognizes California addresses. I didn’t try latitude/longitude.
There are so many inputs that can change a forecast that it’s really hard to account for each factor. Weather obviously is hard to predict at a specific address. But design factors can complicate things too. I have 70 panels. Some from 2010, 2022 and 2023. Had some from 2012 too, but they were swapped. There are 4 different arrays, with 3 different brands of panels. They use the same micro-inverter brand, but there are 2 models in use. The DC to AC ratio is not the same for each of the 4 arrays. And then there is age of the panels, orientation, tilt, etc. Sometimes you don’t have all the information available.
I found that the NREL PVWatts site does a pretty good estimation on a monthly basis. Certainly much more accurate than Solcast ever was. It takes a lot more inputs, too, and many users might have a hard time if an integration required all those inputs.
Personally, I’ll survive without a forecast extension. If somebody came up with a PVWatts-based integration, and then made daily adjustments for weather, I would certainly use it, though.