Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla app version 4.0.0 has app re design for both car and energy

This site may earn commission on affiliate links.
Are you saying it is a physical constraint not to be able to prioritize solar generation to the house and the excess to the grid if the Powerwall is not discharging, or is it a software constraint?
The opposite. Absent a Powerwall (not present or not discharging), the solar generation will always go to the house and then to the grid.

Physically/electrically it can't be any other way than that as the house is an energy sink, the solar is an energy source and the grid can be both an energy source and energy sink. If you could "force" the electrons from solar to go to the grid through the meter then the grid would just have to return electrons back through the meter to the house load.
 
Wow. This is terrible. Now I am wondering if it's a display issue or actually logic. My guess is that the app just informs the Tesla servers what your desire is and they in turn direct the Gateway (PWs) to perform accordingly.

So they either have two separate back end applications based on what client you are on or all of our systems (v3 and V4) are behaving the same way, we just see if differently.

Can you go out to your meter next time you think you have grid load when you shouldn't and verify what the meter says vs what the Tesla app says? That would at least give you one point of reference.
I would be extremely surprised that a Tesla server has any direct control over the behavior of the Powerwall/Gateway as this would create issues when either there is an issue with the server farm, the house's electrical grid, or the network communication from the house to the servers. The Powerwalls/Gateways are operating independently, but they will take updates from those servers (Stormwatch, VPP, user config updates, etc.) and send back data (current flows, voltages, frequencies, state of charge, etc.)

The behavior that @powerwall88 is describing is very likely due the recent changes for Time Based Control with the utilities. The "machine learning" algorithm thinks that it is knows better than the system owner when it is most cost efficient to discharge and it is just wrong (1). The Balanced mode also would discharge in the morning hours powering the house loads and this is similar behavior. It may or may not be taking into account the PW state of charge and how much excess solar is expected for the day. It very likely is not taking into account the complexities of Non-Bypassable Charges (NBCs).

Note 1: Maybe the algorithm is right, but without any details of why it makes sense to discharge the Powerwalls during Off-Peak I stand behind my assertion that the algorithm is wrong.
 
I would be extremely surprised that a Tesla server has any direct control over the behavior of the Powerwall/Gateway as this would create issues when either there is an issue with the server farm, the house's electrical grid, or the network communication from the house to the servers. The Powerwalls/Gateways are operating independently, but they will take updates from those servers (Stormwatch, VPP, user config updates, etc.) and send back data (current flows, voltages, frequencies, state of charge, etc.)
I think we are in agreement here. I did not intend to imply that they operated directly from the server but that they received their instructions from the server. I meant that the apps were not directly communicating with the GW. Therefore there COULD BE two separate server versions based on what app version you were using. This would especially be true with the new features in V4 that don't exist in the V3 app.

Note 1: Maybe the algorithm is right, but without any details of why it makes sense to discharge the Powerwalls during Off-Peak I stand behind my assertion that the algorithm is wrong.
Off peak it should never discharge as you would loose 10% in any scenario to follow. Partial peak I have seen them discharge, but usually only when the AI thinks it will have enough charge to ride out the peak periods before a TOU shift like an upcoming weekend with a larger off peak period.

EDIT: Just saw @Gwgan post while I was typing to describe why off peak could discharge.
 
The off-peak battery discharge may be to allow room to store solar the following day to be sure there is enough to cover the peak periods.
This reason is totally illogical. For most situations, discharging during Off-Peak has no upside for the user. It's a waste of battery cycles. Sure, the PW will get full earlier in the Off-Peak, but it's better to just go into Standby sooner and let the surplus go to the grid at that time.

One could argue that reducing the grid draw during night time hours is better for grid decarbonization, but that is a rather nebulous benefit that has no economic benefit to the user.
 
Can you go out to your meter next time you think you have grid load when you shouldn't and verify what the meter says vs what the Tesla app says? That would at least give you one point of reference.
The grid draw amounts are typically small and brief so I’m unlikely to catch this occurrence but will keep an eye out

TBC is described by Tesla as the mode to use to derive the most economic benefit from PWs, so I agree with @miimura that any off peak discharge typically won’t provide any economic benefit to the owner

I also can’t conceive what rationale can explain solar going to the house during peak, rather than all to the grid and powering the house with PW, as it seems clear to me this does not maximize the economic benefit to the owner
 
  • Like
Reactions: EngineerType
The grid draw amounts are typically small and brief so I’m unlikely to catch this occurrence but will keep an eye out

TBC is described by Tesla as the mode to use to derive the most economic benefit from PWs, so I agree with @miimura that any off peak discharge typically won’t provide any economic benefit to the owner

I also can’t conceive what rationale can explain solar going to the house during peak, rather than all to the grid and powering the house with PW, as it seems clear to me this does not maximize the economic benefit to the owner
Solar going to the house is beneficial if you don't have enough battery energy to make it all the way through the Peak period. However, if you have full net metering without NBCs, then it basically works out the same because you run out earlier, but you got credit for the Peak export.

Now that Tesla has provided the ability to enter the buy and sell prices for each rate period, it should be able to work out the "correct" behavior for any specific situation. If you want to bias the behavior, you should be able to skew the buy/sell prices to get the behavior you want. There was just an example of this posted in the FB group this morning where a guy in Arizona didn't like the AM Peak export because he is on a demand metered plan and he could run out of energy before the end of the morning Peak. Adjusting the Sell price down got him the result he wanted - allowing the solar to go to the house and charge the PWs while maintaining zero grid draw and zero export.
 
I’m never in a situation where I’m at risk of running out of battery prior to end of peak. But I will try playing with the pricing - if I increase the sell price so that it’s significantly higher than the buy price during peak, that should force the system to send solar to the grid rather than the house, right?
 
I’m never in a situation where I’m at risk of running out of battery prior to end of peak. But I will try playing with the pricing - if I increase the sell price so that it’s significantly higher than the buy price during peak, that should force the system to send solar to the grid rather than the house, right?
In theory, yes. However, it should already send the solar to the grid during Peak.
 
The grid draw amounts are typically small and brief so I’m unlikely to catch this occurrence but will keep an eye out

TBC is described by Tesla as the mode to use to derive the most economic benefit from PWs, so I agree with @miimura that any off peak discharge typically won’t provide any economic benefit to the owner

I also can’t conceive what rationale can explain solar going to the house during peak, rather than all to the grid and powering the house with PW, as it seems clear to me this does not maximize the economic benefit to the owner
It isn't much of an economic benefit if you are a net producer. PG&E only compensates you at ~$0.03/kWh for excess production. But you have about a 10% loss when using the Powerwalls. I would rather forfeit the $0.03/kWh and reduce cycles on the Powerwalls.
 
Solar going to the house is beneficial if you don't have enough battery energy to make it all the way through the Peak period. However, if you have full net metering without NBCs, then it basically works out the same because you run out earlier, but you got credit for the Peak export.

Now that Tesla has provided the ability to enter the buy and sell prices for each rate period, it should be able to work out the "correct" behavior for any specific situation. If you want to bias the behavior, you should be able to skew the buy/sell prices to get the behavior you want. There was just an example of this posted in the FB group this morning where a guy in Arizona didn't like the AM Peak export because he is on a demand metered plan and he could run out of energy before the end of the morning Peak. Adjusting the Sell price down got him the result he wanted - allowing the solar to go to the house and charge the PWs while maintaining zero grid draw and zero export.
We should not be reliant on the Tesla app being about to "work out the correct behavior" based on incomplete pricing and usage data. There is a huge amount variation on the base TOU plans both within a single utility and across the thousands of utilities world wide. The nuances of the NEM plans for minimum daily charges (MDCs), non-bypassable charges (NBCs), how the export vs import is compensated for, etc. And then you add on top of this if your system will be a net consumer vs a net generator on a monthly basis for some and a yearly basis for others. This is a sisyphian task that will never be complete due to constantly changing landscape.

Lurking Tesla App Developers - Please just stop trying to solve a task that you cannot solve without manually scrubbing through every regulatory tariff filing by worldwide utilities on a monthly basis or a massive amount of detailed input from the user that most won't understand and just give us explicit prioritization control over the behavior of the Solar flow (house vs powerwall vs grid) and Powerwall charge/discharge behavior by period.
 
We should not be reliant on the Tesla app being about to "work out the correct behavior" based on incomplete pricing and usage data. There is a huge amount variation on the base TOU plans both within a single utility and across the thousands of utilities world wide. The nuances of the NEM plans for minimum daily charges (MDCs), non-bypassable charges (NBCs), how the export vs import is compensated for, etc. And then you add on top of this if your system will be a net consumer vs a net generator on a monthly basis for some and a yearly basis for others. This is a sisyphian task that will never be complete due to constantly changing landscape.

Lurking Tesla App Developers - Please just stop trying to solve a task that you cannot solve without manually scrubbing through every regulatory tariff filing by worldwide utilities on a monthly basis or a massive amount of detailed input from the user that most won't understand and just give us explicit prioritization control over the behavior of the Solar flow (house vs powerwall vs grid) and Powerwall charge/discharge behavior by period.

I completely agree with the sentiment; while the Utility Rate Plan is a huge step forward and likely allows for a lot of tailoring it can't handle all circumstances.

I do however wonder whether the massive amount of standardized data that Tesla will end up with here can help influence simplifications of the unnecessary complications of todays rate plans. Given Tesla now knows your rate plan there are likely lots of "what if's" and comparisons they can asses... which will hopefully show that these plans are actually causing grid instability. [note I'm trying to take the positive approach on the big data / big corporation front but sure there are less consumer friendly uses]

For example while I have set my Utility Rate Plan to work for my dual peak demand charge based plan (v a more of a self powered set up I had previously) I'm using the Powerwall less (self powered in 50's v 80's) and creating much more grid volatility (I import much more from grid in morning and export more in afternoon v smaller draws and exports through out the day). It's working exactly as I would have hoped and I don't believe the economic difference to me is material but I'm more grid dependent. With this sort of accumulated data I think the PC would prefer me to operate in the prior way as it puts less strain on them but their complex plan forces me (and my solar / Powerwall system) to behave differently...
 
(moderator note)

I decided to "unsticky" this thread, not to try to stop discussion in it, but because the tesla app 4.0 came out a while ago now and its my belief this thread can rise or sink on its own.

Much of the discussion in the energy section on this app will likely now pivot to specific utility settings, how they behave, what works for people on specific utility plans, how people try to mimic the behavior of the mode that was removed, etc.

@JayClark started up another thread on the topic of sharing utility settings etc and I have stickied that one. You can find it here:


Of course, everyone can continue to share in this thread if they want to as well, I just thought it was a good idea for people to have a specific thread for sharing those settings. With flexibility also can come confusion, so it really is helpful for those of you who enjoy digging into this stuff to share your settings (so we can copy them if we want the behavior you have meticulously crafted) :cool:
 
I do have 4.2.3 on my phone for the Model S and Model S Plaid. The app is sometimes slower to connect to vehicles but when it works, there's some new functionality, such as being able to change charging amps and setting the scheduled charge time. Android doesn't have it yet but they also recently added the ability for iOS phones to view sentry cameras remotely so I'm sure Android will have it eventually.
I downloaded the latest Andriod app (4.2.3) yesterday and it finally supports editing my utility plan. My peak buy rate is $0.16/kwh, but sell rate is $0.04/kwh for all hours. With this setting, my PW now behaves like the old Balanced mode -- no export during peak unless PW is full.
 
  • Like
Reactions: Redhill_qik
Been a hold out for new app v4 until the complaints subsided...have been using v3.xx as I had not actively updated my IOS app on the iPhone. Been doing fine with access and maintaining balanced mode up until today. Now...no longer have access to my solar/PWs within the app for some reason (still access to the car unchanged). Seems like some have had some recent issues with PW access due to ?? server issues or possibly PWs relying on 3G w- v4 app. I have a PW2 and has been hardwired ethernet and backup wifi connection since originally installed.

Do not know if any others that were holding out on app "upgrade" have noticed the same??
 
Been a hold out for new app v4 until the complaints subsided...have been using v3.xx as I had not actively updated my IOS app on the iPhone. Been doing fine with access and maintaining balanced mode up until today. Now...no longer have access to my solar/PWs within the app for some reason (still access to the car unchanged). Seems like some have had some recent issues with PW access due to ?? server issues or possibly PWs relying on 3G w- v4 app. I have a PW2 and has been hardwired ethernet and backup wifi connection since originally installed.

Do not know if any others that were holding out on app "upgrade" have noticed the same??
I just checked after reading your message and confirmed our Powerwalls just disappeared and are no longer visible in 3.x. We did have a brief outage earlier today and I noticed we are on 21.39.1. I can't recall what previous firmware we were on.

Oh well, it was good while it lasted. I just wish they could add some of the useful graphs that they had before v4.x. With the old graphs, I could instantly get all the information I wanted and don't see any way to get the same information with v4.x. I might have to resort to using a 3rd party application to graph the data.