Tesla: we need more control over our PowerWalls

[Ulmo]

Tesla,

Today around 1PM, you can see I turned off my PowerWalls. Do you know why? Because I wanted to use more of my stored electricity during the evening when solar is not shining and use the solar energy available from the grid during the day during non-peak hours rather than fossil fuels they use at night during peak hours, rather than using all of my stored electricity (sans buffer and 50% Lithium Ion average target settings) before the evening even came. You can see how I achieved that:

http://ulmo.solar/powerflow/2018/11/04.png


I turned off the PowerWalls at 1PM, used about 5kWh about half of which was from PG&E and half from my solar panels, and then turned the PowerWalls back on around 2PM after I had used that power, allowing me to keep an additional 2kWh in my PowerWalls. I ran out of PowerWall energy (sans buffer) around midnight. It was still peak evening hours when the 2kWh that I had saved started being used; i.e., if I had not turned off the PowerWalls and forced grid use during the day during my high usage, I would have been using GRID ENERGY IN THE EVENING PEAK, THE ABSOLUTE WORST TIME TO DO SO.

Tesla, you need to give us better software control of our PowerWalls so that I do not need to physically and manually turn the batteries off during high electrical use during the day; that defeats the backup potential of the PowerWalls, AND is a suboptimal method to operate them. Software should allow us to set the PowerWalls to operate in a manner that maximizes shifting of solar energy into peak dark times.

Yes, I can purchase more solar panels and batteries, when I get more money, and I get more roof space to put them on, and your solar roof tiles are inexpensive and available enough that I can have them reroof the front roof that gets morning sunlight, but there's a few reasons that I haven't done that yet (not the least of which you don't offer those roof tiles now at a cost I can afford and I don't have enough for additional PowerWalls right now), so software should be used in the meanwhile to achieve similar results as much as possible.

Sincerely,
Brad Allen

 

[Yonki]

Very frustrated with this as well. I've been trying to use the app to switch between Backup Only during off-peak and partial-peak (ensuring my Powerwall gets all the limited solar I generate) and Advanced/Cost-Saving during peak. Unfortunately half the time my Tesla app doesn't connect so I miss the the switch going into peak and get screwed. Would be great if this was automatic so at least I couldn't get bit by the poor app connectivity.

Also Tesla has promised me since December of last year that I'd be able to charge the Powerwall from the grid...do any of you "old-timers" believe that's ever going to happen?

 

[llngoc]

Why do you ever want to charge from grid if you have solar? Wouldn't this mess up your SGIP or Federal ITC?

Very frustrated with this as well. I've been trying to use the app to switch between Backup Only during off-peak and partial-peak (ensuring my Powerwall gets all the limited solar I generate) and Advanced/Cost-Saving during peak. Unfortunately half the time my Tesla app doesn't connect so I miss the the switch going into peak and get screwed. Would be great if this was automatic so at least I couldn't get bit by the poor app connectivity.

Also Tesla has promised me since December of last year that I'd be able to charge the Powerwall from the grid...do any of you "old-timers" believe that's ever going to happen?

 

[MorrisonHiker]

Why do you ever want to charge from grid if you have solar? Wouldn't this mess up your SGIP or Federal ITC?

In my case, my Powerwalls came from the referral program. Since I don't live in California, I don't get either the SGIP or Federal ITC. I would like to be able to charge from the grid in emergency situations, such as when it's been a cloudy week and heavy snow is expected.

Yes, they recently added Storm Watch but that still doesn't provide enough flexibility. We recently had several cloudy days in a row <gasp!> and my Powerwalls were drained down to the reserve the first day. I tried increasing the reserve but we had so little solar production for several days in a row that I wan't able to charge them back up. Had the power gone out, I would've only had 25% to power things during the outage. If I'd been able to recharge the Powerwalls from the grid, I would've been able to have 100% charge in the Powerwalls and been able to handle a much longer outage. Fortunately the snow didn't amount to much and we didn't lose power but if it had and it remained cloudy, the Powerwalls and solar panels would've been worthless once the 25% reserve had been used up.

 

[MorrisonHiker]

I agree that we need more control of our Powerwall settings. Balanced TBC kind of works well to cover peak periods except it prioritizes sending to the grid even though my Powerwalls aren't fully charged. It is now pushing back to the grid way before peak starts. Later at night, it stops using the Powerwalls at random times and percentages, not even getting down to the reserve percentage most nights. Since I have a large solar system and multiple Powerwalls, Self-powered almost works for me. However, I would like to be able to schedule an hour or two when I'm not self powered so that I can prevent the Teslas from draining the Powerwalls. I would've added more panels to ensure 100% self-production but our utility sets limits on how much solar one can install.

When I work from home, I usually try to charge my car during the day but TBC still pushes to the grid and drains my Powerwalls to power the house in the shoulder period before peak. I have to remember to switch over to Self-powered an hour or so in advance and then my Powerwalls and car(s) will be charged instead of sending everything back to the grid.

 

[Yonki]

Why do you ever want to charge from grid if you have solar? Wouldn't this mess up your SGIP or Federal ITC?

Well, I want to charge from the grid because I can't plant enough solar to cover my total energy needs. Solar generation is about 9-10kWh/day, while usage averages about 30. So charging the Powerwall from the grid is the only way I can avoid using at least some grid energy at peak and partial-peak. Which theoretically would be in both my and PG&E's best interests.

Tesla didn't say anything about charging from the grid affecting my SGIP or Federal ITC.

 

[Ampster]

Tesla didn't say anything about charging from the grid affecting my SGIP or Federal ITC

Did you read the SGIP paperwork that you signed?

 

[wwhitney]

SGIP doesn't care where you charge from. The only requirement is that you discharge sufficiently.

[If you tell SGIP you are only going to charge from solar, and you apply at a time of high demand that causes a lottery, then you get preference in the lottery. This did apply to Step 1, I believe, but hasn't come up since.]

The ITC cares very much where you charge from. If you are going to charge, say, 80% from solar, then you only get to claim 80% of the normal ITC. If that percentage drops below 75% then the ITC goes to zero (the 75% cliff).

Cheers, Wayne

 

[CA-Gbm84]

Tesla,

Today around 1PM, you can see I turned off my PowerWalls. Do you know why? Because I wanted to use more of my stored electricity during the evening when solar is not shining and use the solar energy available from the grid during the day during non-peak hours rather than fossil fuels they use at night during peak hours, rather than using all of my stored electricity (sans buffer and 50% Lithium Ion average target settings) before the evening even came. You can see how I achieved that:

http://ulmo.solar/powerflow/2018/11/04.png
View attachment 349991

I turned off the PowerWalls at 1PM, used about 5kWh about half of which was from PG&E and half from my solar panels, and then turned the PowerWalls back on around 2PM after I had used that power, allowing me to keep an additional 2kWh in my PowerWalls. I ran out of PowerWall energy (sans buffer) around midnight. It was still peak evening hours when the 2kWh that I had saved started being used; i.e., if I had not turned off the PowerWalls and forced grid use during the day during my high usage, I would have been using GRID ENERGY IN THE EVENING PEAK, THE ABSOLUTE WORST TIME TO DO SO.

Tesla, you need to give us better software control of our PowerWalls so that I do not need to physically and manually turn the batteries off during high electrical use during the day; that defeats the backup potential of the PowerWalls, AND is a suboptimal method to operate them. Software should allow us to set the PowerWalls to operate in a manner that maximizes shifting of solar energy into peak dark times.

Yes, I can purchase more solar panels and batteries, when I get more money, and I get more roof space to put them on, and your solar roof tiles are inexpensive and available enough that I can have them reroof the front roof that gets morning sunlight, but there's a few reasons that I haven't done that yet (not the least of which you don't offer those roof tiles now at a cost I can afford and I don't have enough for additional PowerWalls right now), so software should be used in the meanwhile to achieve similar results as much as possible.

Sincerely,
Brad Allen

Can you explain your graph?

My reading is that you turned off your Powerwall while charging between 1pm and 2pm - so actually you reduced the amount of energy Powerwall had to discharge during peak - which is the opposite of what you describe.

 

[Ulmo]

The following issue is an aside to my original post, but while we're talking about these things, I wanted to respond to this point as well.

I would've added more panels to ensure 100% self-production but our utility sets limits on how much solar one can install.

You reminded me of more complex (mostly dynamic desirable) timing issues which I think can also be done but with more companies involved, but isn't as easy as my initial software request.

I'm guessing your limit on grid connected solar size has to do with local grid capacity; theoretically, the inverter and battery backup controllers could coordinate and set up the solar in such a way that they limit output according to specifications of the utility, either via permanent settings, or dynamic settings. I've envisioned setting up a FLIR camera pointed at the transformers the utility has mounted on poles or on the ground to track their heating and prevent overheating; the utility could do similar. This would require proactive coordination between the utility, the inverter software programmers and hardware engineers if the hardware needs more functions, and the battery software programmers and engineers, but I think is well within reach if they all commit to do so; while the current cost effectiveness of doing such new features may be low, so is the current market penetration of solar-storage-grid-connected home systems and it's going to rise a bunch, so I think it would be worthwhile for companies in development to do the work to make this happen. The grid transformer issue is going to get worse, not better, with more electric vehicles on the roads, so we need to solve this now, where "we" is humans; that includes the above 3 company groups plus the customers. This also includes the need to have EV's schedule their charging according to the capabilities of the grid, minimizing transformer damage (an also avoiding catastrophic failures). This is all in addition to whatever harder limits exist on the local grid wire size imposed limits that would also have to be respected by local generation (solar inverters and the gateways that decide how much energy each thing gets). I'll note that as long as the software is running properly, small spikes in grid use (including generation) that "exceed limits" could/would quickly be abated by the software if programmed properly, and total heating would be thus properly managed. Thus, the software would have to have a sense of the wires involved, to do wire heating modeling, and I recommend FLIR to calibrate those estimates by the software, then safety margins applied.

 

[MorrisonHiker]

The following issue is an aside to my original post, but while we're talking about these things, I wanted to respond to this point as well.

You reminded me of more complex (mostly dynamic desirable) timing issues which I think can also be done but with more companies involved, but isn't as easy as my initial software request.

I'm guessing your limit on grid connected solar size has to do with local grid capacity; theoretically, the inverter and battery backup controllers could coordinate and set up the solar in such a way that they limit output according to specifications of the utility, either via permanent settings, or dynamic settings. I've envisioned setting up a FLIR camera pointed at the transformers the utility has mounted on poles or on the ground to track their heating and prevent overheating; the utility could do similar. This would require proactive coordination between the utility, the inverter software programmers and hardware engineers if the hardware needs more functions, and the battery software programmers and engineers, but I think is well within reach if they all commit to do so; while the current cost effectiveness of doing such new features may be low, so is the current market penetration of solar-storage-grid-connected home systems and it's going to rise a bunch, so I think it would be worthwhile for companies in development to do the work to make this happen. The grid transformer issue is going to get worse, not better, with more electric vehicles on the roads, so we need to solve this now, where "we" is humans; that includes the above 3 company groups plus the customers. This also includes the need to have EV's schedule their charging according to the capabilities of the grid, minimizing transformer damage (an also avoiding catastrophic failures). This is all in addition to whatever harder limits exist on the local grid wire size imposed limits that would also have to be respected by local generation (solar inverters and the gateways that decide how much energy each thing gets). I'll note that as long as the software is running properly, small spikes in grid use (including generation) that "exceed limits" could/would quickly be abated by the software if programmed properly, and total heating would be thus properly managed. Thus, the software would have to have a sense of the wires involved, to do wire heating modeling, and I recommend FLIR to calibrate those estimates by the software, then safety margins applied.

The limit is because the utility company doesn't want people installing more solar than they 'need' so they limit customers to 120% of their previous year's usage. If you buy an electric car, you can also get an allocation for that. We were able to 'cover' 141% of our previous year's usage but we now have 3 Teslas and even more usage. I used to always charge one of the cars at work (free solar Level 2 charging) so that car's usage wasn't represented in our monthly bills. Now I am charging that car at home so that has increased our monthly usage by another ~250 kWh per month. We'll either have to wait another year or buy and register another electric car in order to 'prove' we need more solar.

 

[Ulmo]

Can you explain your graph?

My reading is that you turned off your Powerwall while charging between 1pm and 2pm - so actually you reduced the amount of energy Powerwall had to discharge during peak - which is the opposite of what you describe.

I did turn off my PowerWall while it was charging in the high sun of midday around 1PM. This was intentional, since I knew I was going to use more wattage than my solar system puts out for about an hour (a dryer load of about twice my solar midday output, running for about an hour). Less than a minute later, I turned on that heavy load, and if I had not turned off the batteries, they would have been heavily discharging for 75% of the next hour; instead, I used all of my solar power and made up the rest from the grid, which includes the across the street neighbor's solar panels and the solar panels that the utility has on the grid as well as some more efficient fossil fuels (more efficient than the evening fossil fuels). Midday sun at 1PM-2PM Sunday is off-peak in two ways: in one way, it is not the high-peak of evening use. Here is the duck curve that shows the ramp up that CAISO has to contend with, on the day in question, and fairly common, showing copious utility solar available in the day (and doesn't show customer-side solar) at 1PM-2PM to offset the horrible evening ramp, much of which is fossil fuels:



The second way in which 1PM-2PM is off peak is that the time based rate plan I use is off-peak Sundays until 3PM. Here is the link to the rate plan I use: https://www.pge.com/tariffs/assets/pdf/tariffbook/ELEC_SCHEDS_EV (Sch).pdf. And here is a screenshot of the relevent text:



In terms of rate plans, Sunday evening Peak is ~33 cents per kWh in "Winter" (and ~47 cents per kWh in "Summer"), which is when I discharged the battery (avoiding using peak). Off-Peak is ~13 cents per kWh all year. There is also a Partial-Peak period during weekdays. In my example time period of 1PM-2PM, PG&E is not in PEAK time, whether Summer or Winter and whether weekday or weekend. While I can say that PG&E has not yet properly tuned their rate schedules to the above severe evening peak time periods and growing solar provision during the day, in my example, the mismatched rate periods and the actual ground realities were not mismatched from 1PM-2PM and in the evening discharge from 9PM to 11PM. (Maybe a lawyer can tell me whether "between" is legally inclusive even though in English math interpretation it is noninclusive, since the Sunday in question was indeed the first Sunday in November.)

so actually you reduced the amount of energy Powerwall had to discharge during peak - which is the opposite of what you describe.

Wrong; I think you're getting confused by the pre-solar-power-age weekday industrial factory and light industrial business park terminology of mid-day being "peak". These days, evening is peak, and mid-day is off-peak. If look at the following graph, you can see that the fossil fuel use is least during mid-day, even compared to night time when everyone is asleep, and even compared to the entire day:


That was a Sunday, so I'll put a weekday up for comparison, and while the night time did have some less fossil fuel use times compared to Sunday, it was still more fossil fuel use at any time during the night and twilights than during daytime off-peak, and the trend of more solar power being available during the day is going to continue in the coming decades:


When I say "off-peak" and I'm not trying to talk about the bounds of rate plans, what I mean is when the solar power is at its peak and putting the least net demands on storage and fossil fuel generators, i.e., when we humans need to make up the least amount of power, therefore, off-peak in generation needs. In a way, that terminology is unintentionally ignoring the higher amount of use that industry* has during the day because it is heavily offset by the solar, and also unintentionally ignores the amount of grid capacity needed to bring the electrons from the solar collectors to the heavy users, and that could cause a bone of contention. I suppose when we discuss solar offsetting, we should say "off peak net demand", and when we are discussing grid capacity in wires and transformers, we should say "peak grid use", and "grid use" should mean in any direction (to or from electricity users).

* Some industries slow down their factories during weekday work hours to let maintenance crews have good day time working schedules, then ramp up automated and skeleton-crew based production systems after the maintenance crews go home until the next time they come to work. This was true when I worked in a cement factory, and it is also visibly true on railroads I've lived near in California's Central Valley (where I presume they also used the daytime to load rail cars from shipyards, then filled the rails for nighttime and weekend running).

 

[Ulmo]

http://ulmo.solar/powerflow/2018/11/04.png
View attachment 349991

Note: I didn't spend enough time making the graphs handle the time zone change well (Fall Back at 2AM). You can see where it drew three (!) lines there: two for the actual overlapping use, and one for a backtrace from the end of the first time the hour happened to the beginning of the second time the hour happened. I should have had it show 5 hours between midnight and 4am rather than redrawing the same hour twice. I don't know how to make gnuplot do that, and quite frankly, since this is a free-time one user application, I don't see the need to debug that; the underlying database still has correct data; this is simply a display error.

 

[JohnRatsey]

My complaint is that the logic for doing off-peak charging and how much (I'm in UK) when set to TBC Cost Saving is unpredictable. I reckon I could do better if I had a simple control to set the minimum charge level at the end of the off-peak period which I could adjust after checking the weather forecast and taking account of the likely usage during the next day. The battery would then have two simple rules: (i) If SoC at the start of the off-peak period is above the minimum charge level then continue to discharge during the off-peak period until (if) SoC reaches the minimum; and (ii) if SoC at the start of the off-peak period is below the minimum charge then stop discharging and towards the end of the off-peak period charge from mains up to the minimum value.

I've had both the battery charging to 100% during the off-peak to then be followed by a sunny day and also putting in a small charge when a gloomy day was forecast.

 

[MorrisonHiker]

After finally getting down to within 1% of my set reserve last night, my Powerwalls got up to 80% today. I still haven't figured out why it doesn't charge them all the way when on TBC. It could've easily charged the Powerwalls to 100% and started exporting to the grid over 2 hours early...but no...my system started sending all of my solar production to the grid even though peak doesn't start for another 3 hours. Before 1.25.0 and 1.26.0, I think it used to charge the Powerwalls all the way, if possible, before peak started.

If they could just give me a setting to indicate what hours I want to self-power, I'd be happy!

 

[Ulmo]

After finally getting down to within 1% of my set reserve last night, my Powerwalls got up to 80% today. I still haven't figured out why it doesn't charge them all the way when on TBC. It could've easily charged the Powerwalls to 100% and started exporting to the grid over 2 hours early...but no...my system started sending all of my solar production to the grid even though peak doesn't start for another 3 hours. Before 1.25.0 and 1.26.0, I think it used to charge the Powerwalls all the way, if possible, before peak started.

If they could just give me a setting to indicate what hours I want to self-power, I'd be happy!

Ultimately, giving the individuals more control would then also force the utilities to synchronize their rate plans to what available solar power there is and then the customers would naturally shift their sunlight using storage to darker times as needed. Having Tesla force us to use our batteries in a certain way is not only suboptimal, but is ultimately worse for the environment.

 

[miimura]

The problem I have is that the behavior under TBC-Balanced before 1.25 was predictable and logical. Now, it's just doing crazy stuff with no apparent logic. I am now actively trying to contact Powerwall support to get some answers at to why.

 

[MorrisonHiker]

The problem I have is that the behavior under TBC-Balanced before 1.25 was predictable and logical. Now, it's just doing crazy stuff with no apparent logic. I am now actively trying to contact Powerwall support to get some answers at to why.

I suppose there could be some logic behind the scenes that we aren't aware of. If so, it would be nice if they could show us a message indicating that they intentionally aren't charging the PWs to 100% because of xyz or that they are exporting to the grid because of a request from the power company, etc.

As it is now, it just doesn't make any sense when it is charging or discharging the PWs or when it is sending power to the grid.

 

[JohnRatsey]

The problem I have is that the behavior under TBC-Balanced before 1.25 was predictable and logical. Now, it's just doing crazy stuff with no apparent logic. I am now actively trying to contact Powerwall support to get some answers at to why.

I also found that behaviour with TBC Cost Saving was relatively sensible. Fully charging the battery coincided with the installation of 1.26. I then switched back to Balanced for a couple of weeks (conveniently with an adequate amount of solar generation) and only went back to Cost Saving a few days ago. I wonder if 1.26 wiped some memory of past behaviour which is now gradually being restored.

 

[miimura]

I suppose there could be some logic behind the scenes that we aren't aware of. If so, it would be nice if they could show us a message indicating that they intentionally aren't charging the PWs to 100% because of xyz or that they are exporting to the grid because of a request from the power company, etc.

As it is now, it just doesn't make any sense when it is charging or discharging the PWs or when it is sending power to the grid.

That's why I'm trying to discuss it with Powerwall Support. If there is some intentional logic, I want to know what it is. I also want the ability to opt out of the crazy behavior unless I'm being compensated for it. I want Tesla to implement aggregated demand response and I want to get some extra compensation beyond the normal net metering credits when I contribute energy to it. However, I strongly doubt that's what is going on at the moment.