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

[swedge]

random question: is there a financial downside to just dumping everything to the grid right when peak starts and then drawing a bit at peak rates if necessary for the remainder of the peak period? wouldn't what you draw just effectively cancel out what you sent back to the grid earlier, making it a zero sum game?

There is something like a 10% loss from charging to discharging. On NEM2, there is also ~ 3 cents per kWh NBC cost. So what you propose is not terrible, but not zero sum. During VPP it does do as you suggest, but the $2 per kWh is far higher than peak price, and so does maximize the net credit.

In Export Everything, mine discharges a bunch at the beginning of peak, but then slows down the discharge so that it can continue to cover the house draw till the rated drops. It does quite well, getting to my 20% reserve setting just as the rate drops. It did take a few days for it to estimate my usage patterns.

Below is my grid use yesterday. Export Everything, EV2A rates, Time Based Control. From the day before, the SOC started at my 20% reserve, so the house drew from the grid till solar took over. Around noon, solar was enough to charge back to 100%, so the excess solar was exported. Peak began at 4 pm, and PW dumped a bunch, but then continued to run the house till midnight when the rate dropped. As midnight approached, the PW discharged more to the grid and got to 20% at midnight.

 

[swedge]

Can you tell by examining the system behavior whether with that combination of options the system imposes the restriction that Net Export < PV Production? As a running total, not as an instantaneous constraint.

I.e. With the battery charged both from solar and from the grid, that the grid charged portion of the battery is only used for house loads (load time shifting), and any amount that gets exported to the grid comes from the solar charged portion of the battery (production time shifting)?

Cheers, Wayne

You raise an interesting issue.

(First, a clarification. The limitation is on Export, not Net Export. If you import a bunch and then export it all, your net export is zero. The utility will gladly let you import as much as you want, but don't want you to export what you imported from them, banking the price difference.)

It is certainly possible to export more than solar produces. One could, as an extreme, simply open the breakers to the inverters, and hence produce no solar. Grid charging would proceed, and Export everything would export. So one could conceivably export more than solar production. However, check my thinking below and see if you agree that it won't be a problem.

First, PG&E calculates my estimated solar production each month, and will limit the monthly NEM credit to that amount of export. (This calculation is explicit in the detailed monthly, so called "black&white" bill.) So, my hypothetical above would not be a problem, because PG&E would still estimate potential solar production, and allow that much export.

But back to reality, since the grid charging feature arrived around this April, my system has been doing very little grid charging, because the long, mostly sunny days charge easily to 100%. With no grid charging, all export and house consumption is from solar. So daily export can not exceed solar production, even though the solar export may be time-shifted by the PW so it happens during peak period.

Winter will be a different story, as my cloudy day production will sometimes fall short of full charging, and grid charging will make up the difference. The worst case would be zero solar production, with all charging coming from the grid. Staying with this worst case, if we assume all discharge is to the grid, each day the PW will import and export 80% (charges to 100, discharges to my backup reserve of 20%) or 10.8 kWh. In Dec, Jan and Feb, PG&E estimates my solar production will be 10.3, 9.5 and 9.8kWh respectively. So in this worst case, I would need to self-consume from the PW 1kWh or less per day, on average, to get full credit. Since my peak and partial peak hours of consumption typically amounts to ~50%, or 6.75kWh, and comes from the PW, the export limit won't be a problem in the winter either.

Things might be different for folks in different situations. Larger battery would allow more exporting. Smaller solar would allow less. I think in Calif, larger systems, i.e. 3 or more PW are not allowed to export everything anyway. So everyones milage will vary.

Does this sound right to you?

SW

 

[wwhitney]

(First, a clarification. The limitation is on Export, not Net Export. If you import a bunch and then export it all, your net export is zero. The utility will gladly let you import as much as you want, but don't want you to export what you imported from them, banking the price difference.)

Yes, measuring at the meter. I was perhaps ambiguously using "Net Export" to mean "PW Export - House Loads". To comply with NEM, that should be limited to "PW Charging from Solar," on a time integrated basis.

First, PG&E calculates my estimated solar production each month, and will limit the monthly NEM credit to that amount of export. (This calculation is explicit in the detailed monthly, so called "black&white" bill.) So, my hypothetical above would not be a problem, because PG&E would still estimate potential solar production, and allow that much export.

The excess export would still happen, you just wouldn't get any credit from PG&E for it. So you'd be paying them to charge your PW from the grid, and giving them the energy back for free.

I guess my real question is (and perhaps this will only become clear over the winter): have you or anyone else observed a situation in which Export Everything ends Peak period at above Reserve, with the amount above Reserve matching the excess grid charging that the algorithm estimated you would use to cover your house loads that turned out to be less than forecast?

Basically do we have any evidence the algorithm is smart enough to avoid disallowed grid-based TOU arbitrage?

Cheers, Wayne

 

[cwied]

Basically do we have any evidence the algorithm is smart enough to avoid disallowed grid-based TOU arbitrage?

Cheers, Wayne

I have evidence that the algorithm does not do this on an intraday basis. There was one day where I had 22.7 kWh of solar production and 27.2 kWh of grid export. Since the export limitation is enforced on a billing cycle there is still a possibility that the algorithm might do something on a longer term basis, but I would be surprised if it did.
I've now implemented my own automation to turn off exporting from the battery when the day's exports get close to total solar production.

 

[crackers8199]

In Export Everything, mine discharges a bunch at the beginning of peak, but then slows down the discharge so that it can continue to cover the house draw till the rated drops

mine does this only if I set the sell rate to be 3c lower than the buy rate. which is fine...I'm just trying to figure out if on NEM1 there's any reason to bother with that, or just let it dump everything quickly and then draw when it needs to (even if that is during peak).

 

[JJ-Moonstone]

Mine is set to 20% Backup Reserve, Time Based Control, Export Everything, Grid Charging enabled, and PG&E EV2A rate with $0.03 deduction for sell prices for NEM2 NBC.

It waits to let solar charge, and grid charges at ~3.3kW if needed to get to 100% prior to off peak pricing start.

You might try setting the reserve to 100%, as that may make it start to grid charge, perhaps with a few minutes of delay. This may not be a solution to your issue as It would be a nuisance to need to switch the reserve up and down each day, but it is possible for a script to do that automatically.

SW

Setting the reserve at 100% - we are hoping to avoid having to remember to manually switch the reserve up and down every day, as we originally did this for about three months….. hoping that at some point Tesla will enable the ability to set a time for grid charging (for those with grid charging enabled), in the same way that we can choose what time to charge our vehicle. Seems simple but then I’m not a tech.
With our current settings of 70% reserve for grid outages, just after midnight in off-peak the grid does give the battery a little boost to 80%, but then discharges it back to the 70% by morning (for solar charging?), but our problem is frequent rain/fog for winter months and not much solar, along with outages, and we are hoping to be able to use the powerwall to power home during peak hours, and then charge it from the grid straight away after midnight….. dream on . We actually would ideally set battery at 60% reserve as we typically use about 40% during peak-rate hours, but when we do that, we end up with PW at 60% for most of the day, and we like a fuller battery reserve. We had the Powerwalls functioning this way for a few years before the solar was installed (of course we love having solar!), but in this one regard we miss the confidence that having a nice full powerwall gave us during wintertime. Sometimes we set the reserve at 90% for the confidence part, but it seems crazy to be drawing peak-rates from PGE when there might be a solution that solves everything.
Regarding your settings, I notice that we do not have the option to choose to export everything (or not).
Thanks!

 

[UCIBtheG]

In Export Everything, mine discharges a bunch at the beginning of peak, but then slows down the discharge so that it can continue to cover the house draw till the rated drops. It does quite well, getting to my 20% reserve setting just as the rate drops. It did take a few days for it to estimate my usage patterns.

I've had mine set up to Export Everything for at least a week now. It worked as expected (or close to it) for a couple of days, but doesn't any longer. I've edited my schedule so that there is only Peak from 5pm-8pm (weekdays), the rest is Off-Peak. I also changed the sell price for Off Peak to .03 less than the buy price. I do not have Grid Charging enabled, and have my reserve set to 30%. I'm set to use Time-Based Control.

Looking back in the app, on Sunday it charged from the 30% reserve up to 100%, and didn't export anything, which was good because weekends are all off-peak. On Monday, it powered the home loads from 5pm-8pm, but did not export any power, ending at 82% charge. On Tuesday, there was a tiny blip of solar (from 10:35-10:40am) sent to the PW, but it didn't charge otherwise. It then dumped from the PW starting at the 5:00 peak, and was down to the 30% reserve by 6:30 and pulling from the grid. Wednesday, it had a few brief pops of solar (last one from 2:00-2:10) go to the PW, bringing it from 29% to 35%. It then sent that little bit of stored energy to the grid from 5:00-5:10, then pulled from the grid again. Yesterday there was a small chunk of solar from 9:30-10:50 that brought it from 26% to 31%. It exported next to nothing at peak, because it was already just about at the reserve. So far today, there was another blip to bring the PW up to the 30% reserve, and now excess solar is going to home load and the grid.

I don't know why it isn't consistently (in the pre-5:00 off-peak period) sending all solar to fill the PW, and pulling home load from the grid. I feel like it should have been long enough to establish my usage patterns. I tinkered with some settings previously (switching to Self-Powered to force it to charge the PW, as well as editing the Utility Rate Plan to delete a Super Off Peak period from midnight to 6am, and editing the Off Peak sell price. But those changes were a while back now, and I'm not a patient man Anything else I can check or do?

ETA: I'm going to try setting my reserve to 100% today to force it to charge the PW from solar, and will set it back down to 30% before peak. Is it a necessary step to do that for a few days to make it learn what it should be doing?

 

[swedge]

I've had mine set up to Export Everything for at least a week now. It worked as expected (or close to it) for a couple of days, but doesn't any longer. I've edited my schedule so that there is only Peak from 5pm-8pm (weekdays), the rest is Off-Peak. I also changed the sell price for Off Peak to .03 less than the buy price. I do not have Grid Charging enabled, and have my reserve set to 30%. I'm set to use Time-Based Control.

Looking back in the app, on Sunday it charged from the 30% reserve up to 100%, and didn't export anything, which was good because weekends are all off-peak. On Monday, it powered the home loads from 5pm-8pm, but did not export any power, ending at 82% charge. On Tuesday, there was a tiny blip of solar (from 10:35-10:40am) sent to the PW, but it didn't charge otherwise. It then dumped from the PW starting at the 5:00 peak, and was down to the 30% reserve by 6:30 and pulling from the grid. Wednesday, it had a few brief pops of solar (last one from 2:00-2:10) go to the PW, bringing it from 29% to 35%. It then sent that little bit of stored energy to the grid from 5:00-5:10, then pulled from the grid again. Yesterday there was a small chunk of solar from 9:30-10:50 that brought it from 26% to 31%. It exported next to nothing at peak, because it was already just about at the reserve. So far today, there was another blip to bring the PW up to the 30% reserve, and now excess solar is going to home load and the grid.

I don't know why it isn't consistently (in the pre-5:00 off-peak period) sending all solar to fill the PW, and pulling home load from the grid. I feel like it should have been long enough to establish my usage patterns. I tinkered with some settings previously (switching to Self-Powered to force it to charge the PW, as well as editing the Utility Rate Plan to delete a Super Off Peak period from midnight to 6am, and editing the Off Peak sell price. But those changes were a while back now, and I'm not a patient man Anything else I can check or do?

ETA: I'm going to try setting my reserve to 100% today to force it to charge the PW from solar, and will set it back down to 30% before peak. Is it a necessary step to do that for a few days to make it learn what it should be doing?

That is strange! I think mine took a week or so to settle down. But settle down it did, mostly.

I am just speculating here about why yours might be acting oddly. With different price schedules on weekends, it might need several week end days to learn the patterns who knows... Also, it could be that adjusting settings messes up its learning, I just don't know. In any case, you might want to leave it alone and just watch what it does for a few more days.

From your description, I am not sure exactly what the system did. For example, you said on Sunday it didn't export anything. Do you mean the PW didn't discharge to the grid, or neither the solar nor PW exported anything? If the latter, that would suggest that your solar did not produce enough to cover the house consumption, so there was no excess for charging or exporting. Your Wed story suggests the same, that the solar is not sufficient to carry the house load. But in any case, what you are seeing is strange.

As for your thought that solar should be used to charge PW first, I think I agree. But that is not what mine does either. Mine appears to prioritize covering the house load, and charging (or exporting when 100%) with what is left over. This approach (doing more self powering) may reduce both the importing and exporting, and hence reduce the NBC fees. In any case, mine is not always consistent - sometimes it grid charges for while after midnight, for no reason I can see, and then finishes charging from solar later on. Did it look at a weather forecast saying it was going to be cloudy? Who knows! It just sometimes seems to act a bit random. But so far, mine always manages to gets to 100% before the rate goes up, and gets down to 20% (by running the house and exporting) as the rate drops back to off peak.

If yours doesn't settle down soon, you might want to post some screen shots of the energy graphs so we can see better what it is doing.

SW

 

[swedge]

The excess export would still happen, you just wouldn't get any credit from PG&E for it. So you'd be paying them to charge your PW from the grid, and giving them the energy back for free.

In my installation, the export credit is limited to PG&E's estimate, regardless of how much PV production I have. So if there was no production, I would still get credit for any exporting I did, even if that was PW discharge of earlier, cheaper charging from the grid.

I've forgotten the details, but if I recall correctly, per the NEM tariffs, the estimated production method is allowed for small storage systems (<10kW output ) and uses PV Watts (available online) to calculate maximum production based on location, panel specs, panel count, azimuth, slant. I don't know if it uses historical average cloud cover. But I have never had my export credit limited by exceeding their estimate.

Basically do we have any evidence the algorithm is smart enough to avoid disallowed grid-based TOU arbitrage?

I have not seen any evidence that PW knows about the limit on export credit. Since in 1 and 2 PW installations it is based on the estimation method, PW has no way of knowing what the limit is. Arbitrage is distasteful to PG&E, but the NEM tariffs only limit it to the amount that could be achieved by time-shifting all of one's potential solar production. (Or actual production if the system PTO requires a generation meter.)

I did pay a lot of attention to this earlier this year, because there was a delay in PG&E processing PTO for an additional 6 panels I added to my solar earlier. This meant that my export credit limit was set by the old, smaller PV system. So I tracked my exports, and to keep below the limit I would sometimes slow-charge our car during daylight so that we didn't export that solar. Sadly, the new PTO forced us from NEM1 to NEM2, so we now pay NBC's on our exports.

SW

 

[wwhitney]

In my installation, the export credit is limited to PG&E's estimate, regardless of how much PV production I have. So if there was no production, I would still get credit for any exporting I did, even if that was PW discharge of earlier, cheaper charging from the grid.

Right, but only up to the estimation limit. Any exporting you do beyond that limit, because grid charging is turned on, is energy you bought from PG&E and gave back to them for free.

Because the PW behavior seems at best undefined in this configuration, I personally find it weird to do both Grid Charging and Export Everything at once. If one knew that the Grid Charging was only used for consumption time shifting, and that Export Everything was only used to export PV Production, then it would be fine.

Absent those guarantees, it's a gamble as far as whether the PWs will do the right thing. Of course, since other areas have different NEM rules, there's no particular reason to expect Tesla to implement the California specific limits as a default, but it would be nice to see a configuration option for them. I.e. an "Export All PV" option.

Cheers, Wayne

 

[swedge]

I personally find it weird to do both Grid Charging and Export Everything at once. If one knew that the Grid Charging was only used for consumption time shifting, and that Export Everything was only used to export PV Production, then it would be fine.

Right! Before I started doing this, I modeled it. The issue I was testing was the economics of enlarging my solar so as to cover our EV charging. A large system would loose our NEM1 grandfather status, and Export Everything would leverage our solar production a bit better. It turned out that the NBC cost was minimal, and Export Everything helps substantially more. So we added more than 1 kW and hence NEM2. But the model also showed that grid charging won't happen very often or very much, so as I explained before, we will likely never hit the export limit.

Still, your point that it would be a drag to grid charge but not get paid for the re-export. PG&E would love for that to happen. A 100% tax on power you return to them during peak hours. (Sounds like their dream version of NEM3! ). Anyway, it is good to think it through, and to double check the fine print in the monthly detailed bill.

I haven't thought much about it, but I would not be surprised if some system configurations could get near the limit, as I did when my new larger system was limited to my old smaller estimates. But tracking and avoiding the limit was not hard.

I suspect that it would be harder in a system with a required generation meter, NGOM (sp?). Actual production would be a less generous allowance that the estimates. But NGOMs are required in systems with 3 or more PWs. Since PG&E and Tesla don't allow Export Everything in those systems anyway, the issue shouldn't arise for them.

 

[h2ofun]

Right! Before I started doing this, I modeled it. The issue I was testing was the economics of enlarging my solar so as to cover our EV charging. A large system would loose our NEM1 grandfather status, and Export Everything would leverage our solar production a bit better. It turned out that the NBC cost was minimal, and Export Everything helps substantially more. So we added more than 1 kW and hence NEM2. But the model also showed that grid charging won't happen very often or very much, so as I explained before, we will likely never hit the export limit.

Still, your point that it would be a drag to grid charge but not get paid for the re-export. PG&E would love for that to happen. A 100% tax on power you return to them during peak hours. (Sounds like their dream version of NEM3! ). Anyway, it is good to think it through, and to double check the fine print in the monthly detailed bill.

I haven't thought much about it, but I would not be surprised if some system configurations could get near the limit, as I did when my new larger system was limited to my old smaller estimates. But tracking and avoiding the limit was not hard.

I suspect that it would be harder in a system with a required generation meter, NGOM (sp?). Actual production would be a less generous allowance that the estimates. But NGOMs are required in systems with 3 or more PWs. Since PG&E and Tesla don't allow Export Everything in those systems anyway, the issue shouldn't arise for them.

I had 5 pws, 3 on my one gate way and no NGOM. Now with 7, shall see what happens

 

[cwied]

Right, but only up to the estimation limit. Any exporting you do beyond that limit, because grid charging is turned on, is energy you bought from PG&E and gave back to them for free.

Moreover they subtract credits from the highest rate first, so if the extra exporting happens off-peak it'll end up reducing your credits that cycle.

 

[ziptbm]

Now that the Inflation Reduction Act has been signed, is it safe to turn on Grid Charging without any tax implications claiming the ITC? PowerWalls installed from 1/1/2022 through 2032 will now be eligible for the 30% ITC regardless of whether they are installed with solar panels or standalone storage. I would think that removes the requirement that PowerWalls paired with solar panels be exclusively charged from solar. My solar panels and two PowerWalls went into service in February this year. I've left Grid Charging turned off up until now just to be on the safe side. I would like to turn on grid charging on cloudy or stormy days when my PowerWalls might not get up to 100% by the start of TOU peak hours in the afternoon.

@PhxSunsFan did you find your answer to this? And did you happen to talk with a tax professional?

Now that winter is coming and I'm on a TOU plan, I'd like to charge from the grid overnight, especially after low production days. I just don't want to jeopardize my ITC credit since my system was installed earlier this year. I've got an open question to my CPA on this, but we will see.

One thing I did notice on Tesla's Energy Incentive page is that standalone Powerwalls won't qualify until installed in 2023 or after.

 

[swedge]

I've got an open question to my CPA on this, but we will see.

Be sure to ask your CPA about the down side if you do grid charge and somehow IRS becomes aware. I would expect jail and fines are not a risk. And if a pro-rated repayment of the credit applies, it might be tiny. (I.e. total kWh of grid/solar charging over a year, which would be the case if you had two batteries, one 100% solar and ITC, the other pure grid charged and no ITC.) And would the 2023 version of the rule perhaps apply to or dissuade enforcement on earlier installations?

SW

 

[ziptbm]

Be sure to ask your CPA about the down side if you do grid charge and somehow IRS becomes aware. I would expect jail and fines are not a risk. And if a pro-rated repayment of the credit applies, it might be tiny. (I.e. total kWh of grid/solar charging over a year, which would be the case if you had two batteries, one 100% solar and ITC, the other pure grid charged and no ITC.) And would the 2023 version of the rule perhaps apply to or dissuade enforcement on earlier installations?

SW

This topic appears to be WAY over the heads of most CPAs that know little about solar and energy storage - mine has no clue. I've asked Tesla to clarify if their support page and pop up message in app reflect the changes from the Inflation Reduction Act 2022, but I'm not holding my breath for a firm response on this.

 

[swedge]

This topic appears to be WAY over the heads of most CPAs that know little about solar and energy storage

So true! (The story I shared was about withholding employee tax, i.e. massive literature on that subject, and the Tax Guy was a tax lawyer working for a big 5 accounting firm. In contrast, the solar ITC as it applies to storage is a very narrow bit of tax code.)

The various postings on TMC seemed pretty inconclusive for the same reason - no one has any actual experience with IRS enforcement on this issue. Still, you might want to scan those.

I'm in sunny Calif, with high peak kW prices, while you are in Mi probably with very different details, so the following may not help much. So, in case it helps: I did gird charging on cloudy days all last winter and estimated that it saved me less than $100 on my true-up, because there were not many days when solar couldn't get PW up to 100$. As a percentage of my small true-up it was significant, but $100 over the whole year is down in the noise. So here it might not be worth the risk of loosing the ITC, unless one felt the risk was very low.

Again, I did not take the ITC, so for me those few $ were a bit of gravy. This year, Export Everything should be a much larger benefit, maybe $700, and gird charging will allow that credit to accumulate even on those cloudy days. But this year we started charging our EV at home to eat up that credit and keep our true-up a small balance owing.

SW

 

[h2ofun]

So true! (The story I shared was about withholding employee tax, i.e. massive literature on that subject, and the Tax Guy was a tax lawyer working for a big 5 accounting firm. In contrast, the solar ITC as it applies to storage is a very narrow bit of tax code.)

The various postings on TMC seemed pretty inconclusive for the same reason - no one has any actual experience with IRS enforcement on this issue. Still, you might want to scan those.

I'm in sunny Calif, with high peak kW prices, while you are in Mi probably with very different details, so the following may not help much. So, in case it helps: I did gird charging on cloudy days all last winter and estimated that it saved me less than $100 on my true-up, because there were not many days when solar couldn't get PW up to 100$. As a percentage of my small true-up it was significant, but $100 over the whole year is down in the noise. So here it might not be worth the risk of loosing the ITC, unless one felt the risk was very low.

Again, I did not take the ITC, so for me those few $ were a bit of gravy. This year, Export Everything should be a much larger benefit, maybe $700, and gird charging will allow that credit to accumulate even on those cloudy days. But this year we started charging our EV at home to eat up that credit and keep our true-up a small balance owing.

SW

so what does export everything do? Weird I only have this on one gateway

 

[buckets0fun]

so what does export everything do? Weird I only have this on one gateway

It’ll dump a powerwall to grid, quickly.

 

[swedge]

so what does export everything do?

The default export option is Export Solar Only, which means that once the PWs are charged, solar goes to the house and the rest to the grid. Export Everything lets the PW discharge down to the reserve setting, powering the house, of course, but the rest going out to the grid.

It is fairly smart about it, dumping to the grid as soon the export price goes high, but holding back enough to cover the house till the rate drops.

Weird I only have this on one gateway

Yeah, weird. Perhaps PG&E interconnection folks have brain damage from Low Frequency (60 Hz) EMF radiation. ;-)

I think we have discussed this weirdness a bit before, and that Tesla only allows Export Everything on systems with 1 or 2 PW's, due to PG&E treating system export over 10kW differently. Something like that. If I recall correctly, even those "rules" did not explain your limitations. So I'm thinking LF EMF brain damage.

SW