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

New Powerwall Advanced Options [Toggles for charging from and discharging to grid from powerwalls]

This site may earn commission on affiliate links.
Because by changing the settings all the time, you are never giving the algorithm a chance to settle down. Maybe it's programmed to throw out all the historical data whenever you change the settings, who knows? Just set it to whatever should give you the correct behavior, and don't make changes for 3-7 days. Give it a chance to figure things out. Accept it will do it wrong for a couple days.

Cheers, Wayne


I’ve tried to set it and forget it for a few weeks now… the algo isn’t learning much of anything. My powerwalls continue to behave like what @aesculus posted.

1) Powerwalls dump 15kW for a while
2) then the algo throttles export back, but it just becomes erratic and willy-nilly
3) many times, the rapid dump in the beginning causes the batteries to be unable to last through EV2A peak so I take back peak time energy.

To me, It’s make more sense for the batteries to...
export at 7kW for 4 hours flat during peak time no matter what else is happening
vs
export at 15 kW for 1 hour + export at who the hell knows for 3 hours + home draws from the grid near the end because oopsies doopsies.

But I don't get paid to design brilliant algos, so what do I know.

1655397495806.png
 
Last edited:
@holeydonut I'm with you on this. I would be happier if the Powerwalls would do some calculation and run at a constant charge/discharge rate. In addition to your example above, I am charging from solar during Off-Peak before 3pm and it's drawing from the grid to power the house while it's charging the batteries. At some point it becomes clear that the solar will be more than sufficient to fill the batteries to 100% before Part-Peak starts, so why doesn't it switch to charging from Surplus Solar or charge at a constant rate to hit 100% at the end of Off-Peak? The old Balanced Mode used to transition to Surplus Solar charging in the late morning. I'm on NEM 1, so it doesn't really have any financial impact, but it just looks so inelegant.
 
I have no knowledge of Tesla's algorithm. I have however written a few algorithms.

I don't know if this helps, but here is how I look at the discharge behavior. If you discharged the solar directly to the grid, you would get credit for 1kWh at off peak, or mid-peak rates. If instead, you put 1.1kWh into your Powerwall, you will get back 1.0kWh for later use or export.
  1. If the Powerwall discharge algorithm misses discharging 1.0 kWh at peak, it is one peak kWh of missed credit/income, that day, and the Powerwalls start the next day up 1 kWh. (Undershoot)
  2. If the discharge algorithm and your usage align, you end up at your reserve with no carry over. (Ideal case)
  3. If the discharge algorithm overshoots, and your demand draws a peak kWh that is later offset, the total loss is one NBC+any other noncreditable charges, I.e. small additional cost. (Minimal cost overshoot)
  4. If the discharge algorithm overshoots, and your demand draws a peak kWh that is not later offset, the total loss is 1 peak kWh. (Full overshoot cost)
#4 is only really "knowable" at the end of the net metering year.

In the first case, sooner or later, the algorithm will catch on to the fact that the battery is increasingly full, and ought to trend toward case #2. In the third and fourth cases, the algorithm ought to back off on the discharge for the next day(s). A well designed algorithm for the user, ought penalize #3 and #4 more heavily than #1. A well designed algorithm ought to have some sense of your historical usage patterns (daily, weekly, semifixed N day usage pattern, etc...), to average out past usage, and anticipate future usage. The operative word here being "ought". Normal variation in usage is always going to push/pull the algorithm into one bin or the other. How the algorithm designer weights things in the algorithm determines how much weight recent history and recent settings influence current behavior. The choice on weighting isn't simple, but I would bet that most owners have some sort of seven day pattern that is dominant, and that the pattern changes with seasons and weather.

A simple but greedy algorithm would just do the opposite of yesterday, and you would endlessly cycle around the right behavior. A good one would split the difference. A better one would look at yesterday, and compare where it was in the owner's daily cycle, and adjust accordingly. An even better one might look at microclimate history and forecasts for the owners location... in the end, the issue becomes weighting the relative factors.

From the outside, there is no way of knowing how simple or complicated Tesla's black box algorithm is, and therefore how long one should wait after changing settings for behavior to stabilize, or even if it would ever settle into stable behavior for any given user. (Imagine the usage swings for an AirBnB house...) To make things worse, Tesla is probably comparing the algorithm performance over many users, and changing it, while you are also changing things, with real potential to get way off target.

I think that if case #3 is problematic for you, then it is probably best addressed by daily adjustment of the discharge. An algorithm is never going to know that the person behind you in line at the store barfed all over you, and you suddenly are doing extra laundry, heating more water and using the dryer during peak times.

Personally, guessing that the algorithm has some "knowledge" of weeks, I would give the algorithm at least several weeks between changing the settings, and then I would change the settings only by small increments.

YMWV...

All the best,

BG
 
  • Helpful
Reactions: aesculus
I have no knowledge of Tesla's algorithm. I have however written a few algorithms.

I don't know if this helps, but here is how I look at the discharge behavior. If you discharged the solar directly to the grid, you would get credit for 1kWh at off peak, or mid-peak rates. If instead, you put 1.1kWh into your Powerwall, you will get back 1.0kWh for later use or export.
  1. If the Powerwall discharge algorithm misses discharging 1.0 kWh at peak, it is one peak kWh of missed credit/income, that day, and the Powerwalls start the next day up 1 kWh. (Undershoot)
  2. If the discharge algorithm and your usage align, you end up at your reserve with no carry over. (Ideal case)
  3. If the discharge algorithm overshoots, and your demand draws a peak kWh that is later offset, the total loss is one NBC+any other noncreditable charges, I.e. small additional cost. (Minimal cost overshoot)
  4. If the discharge algorithm overshoots, and your demand draws a peak kWh that is not later offset, the total loss is 1 peak kWh. (Full overshoot cost)
#4 is only really "knowable" at the end of the net metering year.

In the first case, sooner or later, the algorithm will catch on to the fact that the battery is increasingly full, and ought to trend toward case #2. In the third and fourth cases, the algorithm ought to back off on the discharge for the next day(s). A well designed algorithm for the user, ought penalize #3 and #4 more heavily than #1. A well designed algorithm ought to have some sense of your historical usage patterns (daily, weekly, semifixed N day usage pattern, etc...), to average out past usage, and anticipate future usage. The operative word here being "ought". Normal variation in usage is always going to push/pull the algorithm into one bin or the other. How the algorithm designer weights things in the algorithm determines how much weight recent history and recent settings influence current behavior. The choice on weighting isn't simple, but I would bet that most owners have some sort of seven day pattern that is dominant, and that the pattern changes with seasons and weather.

A simple but greedy algorithm would just do the opposite of yesterday, and you would endlessly cycle around the right behavior. A good one would split the difference. A better one would look at yesterday, and compare where it was in the owner's daily cycle, and adjust accordingly. An even better one might look at microclimate history and forecasts for the owners location... in the end, the issue becomes weighting the relative factors.

From the outside, there is no way of knowing how simple or complicated Tesla's black box algorithm is, and therefore how long one should wait after changing settings for behavior to stabilize, or even if it would ever settle into stable behavior for any given user. (Imagine the usage swings for an AirBnB house...) To make things worse, Tesla is probably comparing the algorithm performance over many users, and changing it, while you are also changing things, with real potential to get way off target.

I think that if case #3 is problematic for you, then it is probably best addressed by daily adjustment of the discharge. An algorithm is never going to know that the person behind you in line at the store barfed all over you, and you suddenly are doing extra laundry, heating more water and using the dryer during peak times.

Personally, guessing that the algorithm has some "knowledge" of weeks, I would give the algorithm at least several weeks between changing the settings, and then I would change the settings only by small increments.

YMWV...

All the best,

BG



Uh... isn't that like mega-overcomplicating things?

Like why not just use some super simple math and ignore all the weird stuff? The only benefit I can think of where a hyper-fancy-psychic-learning-algo is better than no algo is if the Tesla Powerwalls are better off exporting at 15 kW for 1 hour vs exporting at 7 kW or 6 kW continuously for 4 or 5 hours.

Like take this zero-brain example (I am good at zero-brain stuff since that's all my brain is good for)...
40 kWh of usable capacity
Reserve to 25%
This is 30 kWh
Utility rate plan entered is 4pm to 9pm
That's 5 hours
30 kWh / 5 hours = 6 kW

So regardless if my home is using all 6 kW, importing, or net exporting... just pound out 6 kW from 4 to 9. Like does it stress out the Gateway or batteries or the Grid itself if this happened?

Based on your scenarios, This zero-brain approach would guarantee a condition of neither undershooting or overshooting total kWh export from the ESS. And error for item #3 would be minimal and would literally take zero-brain to achieve. I'm encountering your item #3 anyway even with the algo. Personally, I don't have to worry about item #4 since my solar array is too small. Edit: ok I misread your error #4. This isn't an issue for me personally because the only reason I imported at peak was because I exported too much earlier in the peak time.
 
Last edited:
Uh... isn't that like mega-overcomplicating things?

Like why not just use some super simple math and ignore all the weird stuff? The only benefit I can think of where a hyper-fancy-psychic-learning-algo is better than no algo is if the Tesla Powerwalls are better off exporting at 15 kW for 1 hour vs exporting at 7 kW or 6 kW continuously for 4 or 5 hours.

Like take this zero-brain example (I am good at zero-brain stuff since that's all my brain is good for)...
40 kWh of usable capacity
Reserve to 25%
This is 30 kWh
Utility rate plan entered is 4pm to 9pm
That's 5 hours
30 kWh / 5 hours = 6 kW

So regardless if my home is using all 6 kW, importing, or net exporting... just pound out 6 kW from 4 to 9. Like does it stress out the Gateway or batteries or the Grid itself if this happened?

Based on your scenarios, This zero-brain approach would guarantee a condition of neither undershooting or overshooting total kWh export from the ESS. And error for item #3 would be minimal and would literally take zero-brain to achieve. I'm encountering your item #3 anyway even with the algo. Personally, I don't have to worry about item #4 since my solar array is too small.

If you and your family are absolutely creatures of habit, the "zero brain" version would be great.👍👍 It does however neglect your usage in peak TOU, but it could just use yesterday's peak TOU usage. The algorithm would wake up at the start of peak period every day, do one calculation, set the power discharge and snooze until the next peak period.

For folks whose daily lives are less routine and who don't live in earth homes or caves, or alter their usage occasionally, weather, air quality, a knowledge of holidays would all enable generally better guesses. As a number of the energy monitoring apps have shown, it is possible to gain some knowledge from the size and timing of current draws, and an algorithm might do that and "realize" that the owner is unexpectedly running laundry during peak periods, and adjust the discharge accordingly. "Might" One could go to town with the algorithm, adding in lunar holidays...To me it all boils down to marginal cost / marginal benefit. There are going to be some population level learning that could go on, but then there is going to be some binning into subpopulations by usage patterns that could be beneficial (e.g. full or partially electric home, etc.).

I could spin scenarios all day, but I have no idea what Tesla is actually doing, where the algorithm lives (The Gateway? The Cloud? The location makes a difference for possible complexity of the algorithm (CPU & memory) and how often one could update the discharge, and to what extent the algorithm could "learn" from others...)

Tesla does employ some bright AI and machine learning types. Whether they are working on this is another issue, as is whether the team has time/resources to refine and / or improve the algorithm. But I would look over at the full self driving mode, and its failures, as a reminder that perfection is hard to obtain in the real world.

All the best,

BG
 
Their new algo, with Time Based and Export Everything, is a bit weirder than without Export Everything on. For example it totally uses my solar for the home load and exports the remainder until the point this no longer works. Then it keeps my home alive while it's starts exporting my stored solar in the PWs until it applies @holeydonut logic at a rate (my home usage) that gets me to reserve at end of peak. I hope PG&E does not choke on the "burst" of energy it gets over a shorter period of time. 🙂
 
  • Informative
Reactions: BGbreeder
With all the changes that have been made I'd like to see if anyone thinks the following operation is possible. I asked this question previously and the answer was no.

During the hours of 4pm-9pm I'd like to run my house off solar with excess production going to the grid. If there isn't sufficient solar to power my home during this period I want run off the Powerwalls.

During the hours of 9pm to 4pm the next day I want to run my home off solar with excess production recharging the Powerwalls. If there isn't sufficient solar to power my home I want to get my additional power from the grid. If it doesn't look like the Powerwalls will be fully charged by 4 pm then I'd like the algorithm to prioritize all solar to recharging the Powerwalls or recharge from the grid as necessary.

Is it possible to do this now?
 
With all the changes that have been made I'd like to see if anyone thinks the following operation is possible. I asked this question previously and the answer was no.

During the hours of 4pm-9pm I'd like to run my house off solar with excess production going to the grid. If there isn't sufficient solar to power my home during this period I want run off the Powerwalls.

During the hours of 9pm to 4pm the next day I want to run my home off solar with excess production recharging the Powerwalls. If there isn't sufficient solar to power my home I want to get my additional power from the grid. If it doesn't look like the Powerwalls will be fully charged by 4 pm then I'd like the algorithm to prioritize all solar to recharging the Powerwalls or recharge from the grid as necessary.

Is it possible to do this now?
Tesla's Powerwall software definitely does not do all of what you want on its own. You might be able to "hack" it with automation. For example, it sounds like you want the Peak behavior to follow the Self Powered algorithm with the battery full at 4pm. Solar will power the house and excess will go to the grid. When the house demand exceeds solar, the batteries will start to discharge so there is no grid draw. However, if the battery is not full, it will still charge with surplus solar in this mode.

During Off-Peak you could switch to Backup mode. The house will be powered by the grid and all solar will go to battery charging. When the solar starts to exceed home load you could switch to Self-Powered Mode. Then surplus solar would charge the batteries. If it looks like there is insufficient solar generation to fill the batteries by 4pm you could switch to Backup mode or TBC to push all solar into the batteries.

It seems that this is not exactly what you want, but it is a reasonable approximation. The automation could be done with Home Assistant or one of the other solutions that access the Powerwall API.
 
Tesla's Powerwall software definitely does not do all of what you want on its own. You might be able to "hack" it with automation. For example, it sounds like you want the Peak behavior to follow the Self Powered algorithm with the battery full at 4pm. Solar will power the house and excess will go to the grid. When the house demand exceeds solar, the batteries will start to discharge so there is no grid draw. However, if the battery is not full, it will still charge with surplus solar in this mode.

During Off-Peak you could switch to Backup mode. The house will be powered by the grid and all solar will go to battery charging. When the solar starts to exceed home load you could switch to Self-Powered Mode. Then surplus solar would charge the batteries. If it looks like there is insufficient solar generation to fill the batteries by 4pm you could switch to Backup mode or TBC to push all solar into the batteries.

It seems that this is not exactly what you want, but it is a reasonable approximation. The automation could be done with Home Assistant or one of the other solutions that access the Powerwall API.
I've thought about this a bit as I'm interested in the same behavior. During summer it easy just to keep it in Self Powered mode with plenty of sun. If I were to use automation I would keep it in Self Powered mode and just have the automation adjust the reserve percentage.

At sunrise have the automation increase the reserve as some function of time so that it hits 100% just before the start of peak then decrease it to whatever reserve you are comfortable with at the start of peak. At the end of peak you can either leave the reserve alone or increase it to whatever the SOC the PWs are at until sunrise when you start the process over again.

It wouldn't work perfectly especially if cloudy weather comes in late in the day but it is the simplest thing I could think of.

EDIT:. Sorry I overlooked the part about grid charging so this wouldn't work. Nevermind.
 
Last edited:
Uh... isn't that like mega-overcomplicating things?

Like why not just use some super simple math and ignore all the weird stuff? The only benefit I can think of where a hyper-fancy-psychic-learning-algo is better than no algo is if the Tesla Powerwalls are better off exporting at 15 kW for 1 hour vs exporting at 7 kW or 6 kW continuously for 4 or 5 hours.

Like take this zero-brain example (I am good at zero-brain stuff since that's all my brain is good for)...
40 kWh of usable capacity
Reserve to 25%
This is 30 kWh
Utility rate plan entered is 4pm to 9pm
That's 5 hours
30 kWh / 5 hours = 6 kW

So regardless if my home is using all 6 kW, importing, or net exporting... just pound out 6 kW from 4 to 9. Like does it stress out the Gateway or batteries or the Grid itself if this happened?

Based on your scenarios, This zero-brain approach would guarantee a condition of neither undershooting or overshooting total kWh export from the ESS. And error for item #3 would be minimal and would literally take zero-brain to achieve. I'm encountering your item #3 anyway even with the algo. Personally, I don't have to worry about item #4 since my solar array is too small. Edit: ok I misread your error #4. This isn't an issue for me personally because the only reason I imported at peak was because I exported too much earlier in the peak time.
Welcome to the era of machine learning and AI!

The basic setup is likely as you describe, with ML on top to tune per home as usage/time/costs evolve. No way to ever turn a sophisticated ML black box back into something a human could understand and make sense of.
 
Welcome to the era of machine learning and AI!

The basic setup is likely as you describe, with ML on top to tune per home as usage/time/costs evolve. No way to ever turn a sophisticated ML black box back into something a human could understand and make sense of.
That is why I am collecting data from my system every minute or so. I am using it to build a system that make decisions on things like reserve percentages, etc. using data from system performance, weather forecasts, etc. I suspect the network will outperform Tesla standard settings by just a few percent. But if something is worth doing it is worth overdoing!!
 
  • Love
Reactions: aesculus
Why in the world would they do that when the peak costs are 2x the partial peak costs?
I have seen it do strange things for a few minutes after I make a change to the settings. I usually toggle various settings, like going off grid or Self-Powered for a moment, till it is doing what I expect it should with the settings I want. I suspect that some of the settings take a while to take effect: the app talks to the server, and the server talks to the Gateway and the Gateway tells PW how fast to charge or discharge. Sometimes it just takes a while to get the message. A few times I have made an change and I let it behave wrong, and after maybe 15 minutes it started working right and got with the program, so to speak.

It is difficult to characterize it's behavior, what with it's "learning", price sensitivity, slow responses to changes, and, especially when it costs us $ to watch it doing the wrong thing, which makes us impatient.

As for the way it discharges quickly during the beginning of peak and then slows down, that is not too different from what I did manually. I calculated the maximum SOC that it could have at the start of each hour and still get down to my reserve at the end of peak. In Export Solar, it would power the house, and in Everything it would add to that the grid dump. So I left it in Solar Only, and each hour checked the SOC. If too high, I'd switch to Everything till it was down to the next hour's scheduled max. So the discharge rate popped up and down, trying to get close to reserve by the end of peak. Now that I'm just letting it do it's own thing in Export Everything, it dumps a bunch, then slows down and makes multiple adjustment to the dump rate as 9 pm approaches. It can't know how much I'm going to use in the house, so it guesses a bit high and dumps when needed to stay near it's glide path. Now, if my wife decides to bake cookies in the electric stove at 8 pm, I expect it might get down to reserve early and then re-import some.

One other point, though. If the buy and sell prices are the same, PW will see little cost to exporting too much and then re-importing if it gets to reserve too soon. It does appear to consider price differentials. But Tesla's EV2-A rates are wrong: not current, and NEM2's NBC is not deducted from the sell. So, unless you adjust the prices, it may discharge too much too soon.

SW
 
  • Helpful
Reactions: aesculus
Hi all I apologize if this has been answered before - I am with PG&E and had my Powerwall's installed through SGIP. According to Tesla support, they are unable to activate grid charging due to my Powerwall's being installed by a third party (Swell Energy). Has anyone else run into this? I originally called Swell Energy who pointed me to PG&E, who in turn pointed me to Telsa. Any help would be appreciated.
Nexb,

I am likewise with PG&E, EV2-A, SGIP, and 3'rd party installer, Semper Solaris in my case. Initially last month, I got the grid charging option, but export control was missing. Every Tesla support person I spoke with had a different excuse, but eventually export control appeared in my app. Sadly, I don't know which call or email I sent got it done, or if Tesla corrected some error in their server software which fixed many customers PWs at once, so I can't tell you who's buttons to push.

If you have one Powerwall, you almost certainly do qualify for these features, Tesla is just confused. Pester 'em till you get someone who agrees with what you want and knows how to make it so.

However, I do think Tesla and PG&E have an issue if you have 3 Powerwalls. These exceed NEM's definition of small systems, and may require a no-export setting on the storage. Likewise if they put you on a multi-tariff plan without an NGOM which meters and reports your solar production. h2ofun has 2 gateways, one of which got the features, and the other not, go figure. I'm not sure even he knows why, but his solar is so large that export everything would not help him anyway.

SW
 
One other point, though. If the buy and sell prices are the same, PW will see little cost to exporting too much and then re-importing if it gets to reserve too soon. It does appear to consider price differentials. But Tesla's EV2-A rates are wrong: not current, and NEM2's NBC is not deducted from the sell. So, unless you adjust the prices, it may discharge too much too soon.
Yeah. They don't have an EV1A, but that was easy to modify the EV2 tables. And I brought down the sell numbers to reflect NBCs too.

Set for another test today with reserve set at 40% from 50%, but had a power outage yesterday so my test was moot.
 
Nexb,

I am likewise with PG&E, EV2-A, SGIP, and 3'rd party installer, Semper Solaris in my case. Initially last month, I got the grid charging option, but export control was missing. Every Tesla support person I spoke with had a different excuse, but eventually export control appeared in my app. Sadly, I don't know which call or email I sent got it done, or if Tesla corrected some error in their server software which fixed many customers PWs at once, so I can't tell you who's buttons to push.

If you have one Powerwall, you almost certainly do qualify for these features, Tesla is just confused. Pester 'em till you get someone who agrees with what you want and knows how to make it so.

However, I do think Tesla and PG&E have an issue if you have 3 Powerwalls. These exceed NEM's definition of small systems, and may require a no-export setting on the storage. Likewise if they put you on a multi-tariff plan without an NGOM which meters and reports your solar production. h2ofun has 2 gateways, one of which got the features, and the other not, go figure. I'm not sure even he knows why, but his solar is so large that export everything would not help him anyway.

SW
Yep, still have the GW with 2 PW's has export, one with 3 does not.
 
  • Informative
Reactions: BGbreeder
Now that I'm just letting it do it's own thing in Export Everything, it dumps a bunch, then slows down and makes multiple adjustment to the dump rate as 9 pm approaches. It can't know how much I'm going to use in the house, so it guesses a bit high and dumps when needed to stay near it's glide path.
I see that now that I am down to 40% reserve. I have two large discharges separated by coasting. I suspect when I drop to 30% I will have three large discharges and three coasting periods.
Now, if my wife decides to bake cookies in the electric stove at 8 pm, I expect it might get down to reserve early and then re-import some.
I suspect if she baked the cookies earlier they would get folded into the dump and cruise algo, but at 8PM, you are on your last cruise and it would have to go to the grid to make up the difference. Will that be a new rule in your family to always bake more than an hour before dropping out of peak period? 🙂
 
Finally I have some estimates of the economic impacts of the new features. We are adding some solar and trying to size it to cover the cost of charging our Tesla Model Y.

First, two caveats:
  1. Every system is different, so your milage will differ.
  2. These are rough estimates and a complicated calculation, they are not trustworthy anyway.
I will describe my system and how I estimated below, but here is the bottom line:

Worst case, solar but no NEM, no export. Total annual electrical cost: ~$1,100.​
NEM2, Solar, Powerwall in Time Based, Export Solar Only mode: ~ $ 580. This saves me ~ $520 per year.​
The same but PowerWall in Export Everything mode. Total annual electrical cost: ~$190. This mode saves another $390 per year.​
Grid charging will save less than $100 more.​

So, Export Everything will be my setting going forward, 20% reserve, and will save me nearly $1,000 at true-up. I'll keep my fingers crossed all year for the anticipated minimal true-up.

One other observation. It is cool that PW gives this extra economic benefit to solar, but that alone would not justify the PW cost. Resiliency during power outages is hard to put a price on, but SGIP thankfully made these issues moot for us. It was amazing while it lasted!

The system is 6kW DC Solar, Enphase micro-inverters, one PowerWall. We are on California's PG&E EV2-A TOU rate plan. I estimated the car use as 16,000 miles at 0.32 kWh/mi based on 15 months of charging history thanks to TeslaFi. (Interestingly, the car uses about as much as the house does.) Hourly house consumption (~14kWh/day) is based on 2021 data from Enphase consumption monitoring. Likewise solar production is from Enphase, with a 55% boost for the new solar addition. The calculation was monthly and involved a fair amount of averaging and interpolation. To simplify, I broke consumption and production into 6 daily time periods which correspond to PW behavior modes: night, sunup (PW charging), PW full (off peak), partial peak (starting 3 pm), peak 4 pm, and partial peak evening 9pm. Excel's non-non-existent documentation and column/row naming of variables means errors are far easier to make than valid models are, so I could be way off. But it did give me import and export numbers very close to PG&E's for 2021, in spite of calendar month vs billing month offsets. Sorry, but the "code" is way too ugly to share.

The bad news from all this for me is that adding more solar means I will loose my NEM1 grandfathering and PG&E will extract a few more bucks from me each month when I give them power. Hopefully I can get it done before NEM3 comes along.
 
  • Informative
  • Like
Reactions: h2ofun and aesculus
Finally I have some estimates of the economic impacts of the new features. We are adding some solar and trying to size it to cover the cost of charging our Tesla Model Y.
Just for the heck it: How does your calcs align with the Impact value Tesla calculates in the app? It appears to be the "sold back" value, and not the total net costs/payment.

I did not see a way to download it, but if you did it month by month it might not be too bad manually.

I am coming up for my TrueUp next month and will attack it then. Graphically I cannot see a huge difference for the week I have run Export Everything and I also only had about 10-15 kWh to give back which might be $5-$10 a day extra at best. The weather (cloud cover and heat) now has a bigger impact on my production vs the month of May when we had mostly clear skies and lower temperatures.
 
How does your calcs align with the Impact value Tesla calculates in the app?
Well.... my app is still not showing the Value card in the Impact page.

$5 per day, $35 per week, $150 pre month, $1,800 per year. Not bad. Of course you might need to do some grid charging on short or cloudy days to get to 100% SOC prior to peak.

One thing that had me troubled was what if I just added 2kW more solar and didn't tell PG&E. Sadly, my model says I would export more than their estimate, and so loose peak credit for the "excess" export. The same could happen with grid charging, which of course is why they limit how much export they'll credit you for. Of course, any direct consumption of solar production give you margin, and they only look at the monthly total, so a few days of grid top-off, even in winter, probably wouldn't break the limit. Still, if you do some grid charging, it might be best to check the PG&E estimate so you don't go over. I charged my car some during the day this month for this reason. Heck, that also saves the NBC fee on re-importing too, though loosing a $0.55731 and a $0.02449 fee are a different lever of concern...

I do need to pester Tesla again about that value card. They sure do seem to have a lot of ways to screw up each customer's app differently...