1.46 FW and Standby Mode

[aesculus]

Last night my PWs got 1.46 of the firmware. I have been running in Cost Based mode at 50% reserve for months and except for one event last week all seemed to be good. PW will charge to 100% during the day and flip to discharge at peak (3pm). Generally finish peak around 75%.

Today the powerflow shows nothing going to PW at 10:30AM and all the solar going the grid and house. PW are only at 75% SoC. If I flip the settings to backup mode it immediately diverts solar to the PW and uses the grid to power the house. Flipping it back to Cost Based mode again sends all the solar power to the grid and house. And even worse starts bleeding off what I put in the PWs to the house. And the numbers don't add up either. See picture.

Is there a bug in 1.46 or does my system have to relearn something? I suspect if I don't override this Standby mode that by peak I will not have sufficient SoC to carry me through until the end of peak period.

 

[cwied]

What rate plan are you on? I think that in cost saving mode, the Powerwalls may try to discharge to the reserve during shoulder periods by the end of a Friday if there is cheap power available on Saturday. That's the case with the PG&E E-6 rate plan, for example.

 

[dortor4ev]

I’ve noted that it’s someone non-deterministic as to when the powerwalls’s will be charged in any mode other than ‘self powered’ - I’ve seen entire mornings of excess solar sent to the grid while the battery remains less than fully charged - and then late in the afternoon it starts charging from solar for no apparent reason

I know the software does historically modeling and trys to predict when the best time to charge the battery is - but I believe the algorythms are “flawed” in that while historical usage can indicate the best time to do something, you can’t predict when you’ll have excess solar due to solar variability (cloud cover anyone?? dropping solar production right when it would’ve charged the battery)

I’ve taken to to “forcing” the powerwalls to charge by by switching to “self powered” when my solar is rocking the situation, and then switching back once the batteries are at 100%

but yeah - the data upon which the powerwalls choose to charge form solar is not represented in the Tesla app - and therefore the powerwalls appear to not charge even when one would think “now is a good time to direct some solar power to the batteries” - i.e. use it while you have ahve , because it might be gone in a few moments…

 

[aesculus]

What rate plan are you on? I think that in cost saving mode, the Powerwalls may try to discharge to the reserve during shoulder periods by the end of a Friday if there is cheap power available on Saturday. That's the case with the PG&E E-6 rate plan, for example.

I think you hit the nail on the head. On my post above I stated I had called Tesla with a similar issue (wait for it) last Friday, but I had forgot the details. They just told me it uses a weekly average and that issue (of not charging above 80%) sounded acceptable.

It would have been good if they could have told me the same thing you sumized, that Friday it will limit it's powerwall charging because Saturday is a whole new ballgame.

For my EVA rate plan Saturdays peak is 3-7 and there is no shoulders so it does not really need to store up too much energy for Sat or Sun.

 

[zanary]

I've also noticed the same behavior and what I've come to the concussion is that that behavior is consistent week by week. I did a test not too long ago, and changed the weekend Price Schedule to be the same as weekday, and that also changed the behavior.

I have a feeling that on Thursday and Friday it behaves differently because over the weekend for me, most of the solar generation goes into the Powerwall, so Thursday and Friday it choose to discharge battery to run my house and tries to NEM all generation back.