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

[holeydonut]

@Vines

Do you know if the Tesla Powerwalls are better off exporting at max kW over a short duration, or exporting at 3 kW over a long duration? Like have you seen any advice/whitepaper stuff from Tesla measuring the impact to energy density or the on-board inverter or the cooling module?

For my system with 3x Powerwalls and about 39 kWh total. If I could reasonably predict that the 3x Powerwalls would drain down to 20% during the 9-hours-long EV2A shoulder + peak time (using 30 kWh so I'm left with 9 kWh)... would I be better off just having the batteries dump 30 kWh from 4pm to 6pm (15 kW x 2 hours)? Or does this stress the hell out of the Powerwalls if I did this every day?

 

[Vines]

@Vines

Do you know if the Tesla Powerwalls are better off exporting at max kW over a short duration, or exporting at 3 kW over a long duration? Like have you seen any advice/whitepaper stuff from Tesla measuring the impact to energy density or the on-board inverter or the cooling module?

For my system with 3x Powerwalls and about 39 kWh total. If I could reasonably predict that the 3x Powerwalls would drain down to 20% during the 9-hours-long EV2A shoulder + peak time (using 30 kWh so I'm left with 9 kWh)... would I be better off just having the batteries dump 30 kWh from 4pm to 6pm (15 kW x 2 hours)? Or does this stress the hell out of the Powerwalls if I did this every day?

The Powerwalls would be better off at a lower discharge rate I think. I don't have anything to back this up, but it makes sense that driving anything significantly below its limit will make it last longer.

Also it is a known feature of batteries that the slower you discharge them the more you can get out of them. This is why manufacturers typically rate cell capacity at a certain discharge rate. This rate is measured in C or total pack capacity. So a Powerwall is rated for approximately a discharge rate of 0.43C

Meanwhile, a Model 3 performance has approximately a 5.5 C discharge rate for short durations. Ask me how I know how much faster the electrons go at 413 kW of output!

Meanwhile, if you took only 100w out of a Powerwall constantly for hours it might last like 140 hours instead of the expected 135 hours, since batteries hold more charge when you discharge them gently.

 

[MJ_CA_2019]

Man, this math is complicated...

What is the best way to arrange it so that I can intelligently try to send a total kWh back to the grid at peak time per day? Since my solar is undersized, I'm at an annual kWh deficit of about 1,500 kWh per year due to my wife's Tesla.

With the price-gap between EV2A peak/off-peak and a 8% efficiency lost, I feel like I need to push 1,500 kWh extra to the grid at EV2A peak time during the "Summer" (PG&E says that's June 1 to Sept 30).

That's 12.5 kWh per day I need to push on average to the grid during these Summer 120 days to compensate for PG&E having blocked me from adding 2-3 extra solar panels. But how do I intelligently set it up so I push 12.5 kWh? The all-or-nothing approach on the app now just exports until the reserve is hit, which is problematic since the batteries stop exporting completely and my home will take power back from the grid under EV2-A peak time.

It's almost like I need a double reserve... a first reserve for this "export everything" and a sub-reserve where the batteries will only discharge down to this sub-level to power home loads under the normal time-based-control.

If the 8% conversion losses weren't an issue, then it wouldn't matter if my batteries just discharged down to the reserve every day. But losing 8% of the Powerwalls each day is worth almost 350 kWh over the summer 120 days.

Edit, but now that I think about it, if I export at peak and then take power back from PG&E a few hours later in that day at the shoulder rate, technically I come out ahead? Maybe? AAaaahhhhh @miimura and @Redhill_qik set me straightttttt. Should I just "export everything" down to a reasonable reserve during EV2A summer and just say eff it?

How about this?

Modify the prices in the rate plan so that the system thinks it isn't worth it to run your home off the battery in the "early shoulder" period.
This would be during PV generation hours, so once your PW are topped off, the system would export any excess. Downside is it would run the house off grid power if PV doesn't meet your home demand, even after PW are full.

Are your 3 PW enough to get through the "peak" and the "late shoulder" period?

 

[MJ_CA_2019]

My solar system is too small, I don't actually generate enough kWh each year compared to my annual usage. Unlike you, PG&E blocked me from going over 110% my previous year usage even though I was getting an EV and there were more people planned in the household. Also, my NEM2-MT agreement only lets me export from my batteries, I cannot grid charge.

So what I need to do is export power during Summer-peak time (starts June 1 and ends September 30) to generate credits at peak rates.

This will then let me take energy back from PG&E in the Winter-off-peak or even Winter-shoulder/peak time since I can't grid charge in the Winter. The Summer-peak EV2A rate is almost double the off-peak rate, so more than makes up for the ~8% round trip loss.

Don't get me wrong, it'd be awesome if I could both grid charge and grid export... I'm just not allowed to grid charge. As Wwhitney said at the start of this thread... you're supposed to have one or the other; not both.

View attachment 811078

I think we have talked about this before, but this it my Appendix I. I'm also on NEM-MT, was told I can't be on EV2A because NEM-PS is required. *shrug*
I don't have the "Solar vs Export All" option in my Tesla App. I do have the Grid Charge option.

 

[holeydonut]

I think we have talked about this before, but this it my Appendix I. I'm also on NEM-MT, was told I can't be on EV2A because NEM-PS is required. *shrug*
I don't have the "Solar vs Export All" option in my Tesla App. I do have the Grid Charge option.
View attachment 811659

We need a sticky thread where we all upload our interconnection agreements so we can compare/contrast how effed up they all are.

I tried to check that second box when I got PTO'd and PG&E came back and denied my PTO saying I had to check the first box lololol. At the time, I said Stormwatch may not conform with that checkbox, so I wanted to be like you with the second box checked. This sent PG&E and Sunrun into a 3 week long investigation with Tesla over whether this meant Tesla would have to disable Stormwatch permanently on my Powerwalls lololol.

In the Summer when it's hot enough to run AC a lot... I can go from 3pm to midnight (the entirety of shoulder and peak) with my 3x Powerwalls. And usually end with 25% reserve by 12:01am. The exceptions were those brutal 110F days where the ACs never turned off, and I required 100% of all 3 Powerwalls to midnight.

BTW regarding what Vines said about the impact to the batteries each dumping at 5 kW over the peak time... I think at this time I'm going to have to re-think my strategy of using some peak-time shifting to generate excess NEM credits. I don't think depleting the longevity of the Powerwalls is worth the ~$300 extra credits I was trying to generate. Assuming Powerwall capacity drops 20% over their useful life even under normal use, I feel like frying them to do peak shifting just isn't worth it.

I really wish those effers just let me put the 3 goddamn extra solar panels on my flippin roof. Those 3 would get me about 1,500 kWh over a year and that'd get my wife 5,000 miles in her Performance 3. If ChargeHQ works as advertised, that energy would flow straight from my roof into the car, and not even "burden the fixed costs of PG&E's grid". And all those jerks who say charging a car with "dirty grid overnight energy" could just shove it since it'd be 100% solar. Easy peasy. Barf.

 

[cwied]

BTW regarding what Vines said about the impact to the batteries each dumping at 5 kW over the peak time... I think at this time I'm going to have to re-think my strategy of using some peak-time shifting to generate excess NEM credits. I don't think depleting the longevity of the Powerwalls is worth the ~$300 extra credits I was trying to generate. Assuming Powerwall capacity drops 20% over their useful life even under normal use, I feel like frying them to do peak shifting just isn't worth it.

One thing to consider is that compared to ev batteries, home batteries hardly have any stress. 5 kW per Powerwall is only .37 C (1 C is one charge or discharge per hour). Here's an early blog entry from Tesla mentioning that it's charge rates higher than .5 C that affect battery life: A Bit About Batteries

Note that the fastest Tesla cars can hit a discharge rate of greater than 4 C, though they wouldn't do this for very long.

It seems to me that Powerwalls are probably sized to comfortably support full discharge rate.

A little anecdote that probably doesn't prove anything: since I've turned on export everything, my reported battery capacity has actually gone up. It was about 26900 a month ago, and today is 27900. According to the total export, I'm at about 375 full discharges on my batteries right now.

 

[swedge]

Do you know if the Tesla Powerwalls are better off exporting at max kW over a short duration, or exporting at 3 kW over a long duration?

For the PW 2, Tesla specifies the 13.5 kW capacity and the 90% round trip efficiency at 3.3kW charge and discharge rates. I expect that is probably close to the optimum, or they would have chosen a different rate for the specs.

PowerWall 2 AC Datasheet

 

[swedge]

The all-or-nothing approach on the app now just exports until the reserve is hit, which is problematic since the batteries stop exporting completely and my home will take power back from the grid under EV2-A peak time.

I think you and I are in similar positions, hoping the peak rate credits will leverage our solar enough to cover EV charging so the true-up doesn't sting too much. I am thinking that Export Everything may be a big help, but, like you said, the math is not trivial.

Regarding the problem you cite above, I edited my rate plan to set export 3 cents lower than import on EV2-A, simulating NEM2 NBCs. Now in Export Everything, Time Based Control, my PW throttles back it's discharge so that the charge is down to my 20% reserve at the end of peak at 9pm. After that, the house imports at partial peak, but as you realized, that energy was exported at the higher peak price, so it is still a win.

In the old days, i.e. last month, prior to Export Everything, it was simple: PW would cover our house draw during peak and partial peak, meaning our import was all off peak. Whatever excess solar was exported during peak and partial, well that was gravy, net credit toward our EV charging.

It is more complicated now, and depends on how the PW actually behaves, which we are all still trying to verify. My current thinking is that this new PW behavior may be exactly what we want.

SW

 

[swedge]

Should it blindly discharge from the start at max rate down to the reserve, or just pump out a little more than needed to reach reserve around the same time as end of peak power, or wait until the last 30 mins and pump out at the max speed to reach reserve.

Today, mine let excess Solar export till that was below 3.5 kW, then throttled up to keep the total grid export at 3.5 kw. As the SOC approched my 20% reserve setting, it throttled down to keep the grid flow near zero, discharging to cover house load. I assume it guessed that I would not start baking cookies at 8PM, based on prior days data. It ended the peak period with SOC exactly at my reserve setting.

Of course, your rate plan, situation and setting no doubt differ from mine.

Here are the grid and battery views of today's info:

You can see that onece the solar started fading down, the PW picked up and kept grid export steady. Till SOC around 25% when it slowed down and net flow went to zero till the end of peak.

I am thinking that a second PW would let me do better, but that is a calculation for another day.... For now I am just pleased to see the PW doing what appears to be pretty much the right thing with this new Export Everything feature.

SW

 

[holeydonut]

Today, mine let excess Solar export till that was below 3.5 kW, then throttled up to keep the total grid export at 3.5 kw. As the SOC approched my 20% reserve setting, it throttled down to keep the grid flow near zero, discharging to cover house load. I assume it guessed that I would not start baking cookies at 8PM, based on prior days data. It ended the peak period with SOC exactly at my reserve setting.

Of course, your rate plan, situation and setting no doubt differ from mine.

Here are the grid and battery views of today's info:

You can see that onece the solar started fading down, the PW picked up and kept grid export steady. Till SOC around 25% when it slowed down and net flow went to zero till the end of peak.

View attachment 811697View attachment 811699

I am thinking that a second PW would let me do better, but that is a calculation for another day.... For now I am just pleased to see the PW doing what appears to be pretty much the right thing with this new Export Everything feature.

SW

Man I’m an idiot… my peak time is 4pm to 9pm (6 hours) … not the 3 hours I said earlier. Probably shouldn’t post on TMC during meetings lol.

I’ll set a reserve of 25% and let the Powerwalls do their thing a few times under “export everything” and see if they learn to moderate export the way your charts show over time.

 

[aesculus]

Regarding the problem you cite above, I edited my rate plan to set export 3 cents lower than import on EV2-A, simulating NEM2 NBCs.

A nitty question. NBCs are added on to the rate cost during import? If that is correct wouldn't you add the 3 cents to import vs subtracting from export? I know that the behavior would not change, but it would reflect a more accurate price on the app, correct?

Or are NBCs factored into the rate cost and when you export, you are not credited with that portion?

So confusing.

 

[UCIBtheG]

Does anyone know for sure how Powerwalls (or the associated app) do energy usage prediction? For example, in summertime, does it just look at the trends of the last week/month, or does it look at previous year(s) energy usage (if available)? Does it take the daily weather forecast into account? If a day is going to be a scorcher to where A/C needs to run constantly, might it hold off discharging the PW(s) in partial peak to ensure that enough energy is available to cover the peak period?

FWIW, I have just 1 PW, and for the summer period last year, I set my early partial peak (from noon to 5pm) as off peak so it wouldn't discharge in that period. This year, I have a newly installed whole house fan and attic fan that should help reduce AC usage. I restored the partial peak period in the app, and yesterday (first day of summer rates) the house ran from noon to midnight off the PW, and it was at 66% when it started charging from the grid at midnight. But that was with no AC usage (only got to 95 degrees, and we were out of the house last night). There will be much hotter days ahead that will definitely need copious AC usage. I'm hoping that I won't have to make any manual changes again to ensure I'll be covered for the peak period.

 

[aesculus]

Does anyone know for sure how Powerwalls (or the associated app) do energy usage prediction? For example, in summertime, does it just look at the trends of the last week/month, or does it look at previous year(s) energy usage (if available)? Does it take the daily weather forecast into account? If a day is going to be a scorcher to where A/C needs to run constantly, might it hold off discharging the PW(s) in partial peak to ensure that enough energy is available to cover the peak period?

FWIW, I have just 1 PW, and for the summer period last year, I set my early partial peak (from noon to 5pm) as off peak so it wouldn't discharge in that period. This year, I have a newly installed whole house fan and attic fan that should help reduce AC usage. I restored the partial peak period in the app, and yesterday (first day of summer rates) the house ran from noon to midnight off the PW, and it was at 66% when it started charging from the grid at midnight. But that was with no AC usage (only got to 95 degrees, and we were out of the house last night). There will be much hotter days ahead that will definitely need copious AC usage. I'm hoping that I won't have to make any manual changes again to ensure I'll be covered for the peak period.

I have absolutely no evidence but I think it only looks at the last week or less. Perhaps keeps a view based on different periods (ie we have a weekday and weekend period).

Its easily fooled when weather changes in my experience. So if it clouds up or rains, I have seen it make some big errors. Living where I do, that is generally not a problem as the weather is very stable during prime solar production months.

And of course if you change your behavior like cooking a turkey or changing temp settings when you don't do that daily, it will be off too. For example we had guests for 2 weeks last year. The first week it was way off, the second ok again, the third off again and the fourth OK.

 

[holeydonut]

A nitty question. NBCs are added on to the rate cost during import? If that is correct wouldn't you add the 3 cents to import vs subtracting from export? I know that the behavior would not change, but it would reflect a more accurate price on the app, correct?

Or are NBCs factored into the rate cost and when you export, you are not credited with that portion?

So confusing.

I view things the same as swedge. It's easier for me to conceptualize things as the NBC being a cost to export; which diminishes the export value.

This is because my net imports in kWh from the grid are greater than my net exports. So when I import from the grid for non-NEM kWh, then I want to cost that at the true retail rate on imports. The NBC for imported energy only applies to the NEM portion (the NBC goes up to the limit of your exported kWh).

If you're someone like h2ofun who pushes out megawatts more than he ever uses, it may make more sense for him to just increase his import rate by $0.03 per kWh. But then h2ofun's export value would still be wrong in the app. He's only getting the retail rate o nexports up to his NEM true-up. His export rate drops to like $0.03 per kWh for some of his production since he's a net generator. But there's no way the Tesla app can cost that out.

 

[SoundDaTrumpet]

I have absolutely no evidence but I think it only looks at the last week or less. Perhaps keeps a view based on different periods (ie we have a weekday and weekend period).

I have a weekday and weekend periods as well. The PWs behaved differently that how it was in 2021. My PWs used to charge to 100% SOC, but more recently charged to a lower 50% SOC. The logic appears as if PWs wanted me to reach the reserve each day at the end of peak. This behavior is related to: (a) deeply waiting for the off peak weekends to charge to 100%, and/or (b) keep a lower SOC for battery health (unlikely IMO). / Recently switched to “export everything” and now above doesn’t hold true anymore AFIK.

 

[h2ofun]

I view things the same as swedge. It's easier for me to conceptualize things as the NBC being a cost to export; which diminishes the export value.

This is because my net imports in kWh from the grid are greater than my net exports. So when I import from the grid for non-NEM kWh, then I want to cost that at the true retail rate on imports. The NBC for imported energy only applies to the NEM portion (the NBC goes up to the limit of your exported kWh).

If you're someone like h2ofun who pushes out megawatts more than he ever uses, it may make more sense for him to just increase his import rate by $0.03 per kWh. But then h2ofun's export value would still be wrong in the app. He's only getting the retail rate o nexports up to his NEM true-up. His export rate drops to like $0.03 per kWh for some of his production since he's a net generator. But there's no way the Tesla app can cost that out.

I just leave mine on self powered and never worry about it.

I have a friend coming by tonight so he will be the first Tesla I get to charge for free

 

[holeydonut]

I just leave mine on self powered and never worry about it.

I have a friend coming by tonight so he will be the first Tesla I get to charge for free

Hopefully this guy convinces you EV's aren't tiny and you fit in one just fine lol

 

[aesculus]

OK. Totally still confused on how NBCs are charged. Read the above posts (and many others), read PG&E's laymen's treatise, and walked away confused.

Could someone put it succinctly when I can expect to be charged NBCs and how it's calculated? Also it appears from my bills to be a monthly factor, that gets accumulated at true-up, but not adjusted?

FWIW most months during summer I am a net generator, but never off peak. Most winter months I am a net consumer. With PWs I am always a generator during peak. So basically some months I over produce during partial peak and some months I overconsume during partial peak.

 

[aesculus]

Hopefully this guy convinces you EV's aren't tiny and you fit in one just fine lol

He better be driving an MX.

 

[wwhitney]

Could someone put it succinctly when I can expect to be charged NBCs and how it's calculated? Also it appears from my bills to be a monthly factor, that gets accumulated at true-up, but not adjusted?

I'm on NEM1, so I may have this wrong, but my understanding:

The rate is quoted as something like $0.33/kWh (to make up a number), of which $0.03/kWh is NBC (likewise). So when you import a kWh you get charged $0.30/kWh for energy and $0.03/kWh for NBC. When you export a kWh you get $0.30/kWh credit for energy.

At true up you pay all your NBC charges, plus your net energy charges, if that's a debit balance. If your net energy charges are a credit balance, you lose it, you can't use it against the NBCs (that's the non-bypassable part). And as usual, if your net energy charges are a credit balance, and you are a net kWh exporter, you get a payment per net kWh exported. As a straight payment, that can offset any charges, like your monthly minimum and your NBCs.

Now what I don't know is how NBCs interact with monthly minimum charges. I.e. if there's a $5/month monthly minimum (again to make up a number), or $60/year, and at year end you have $50 of NBCs, with a net credit in energy $ (but a net consumer of kWhs), do you pay $110 or $60 or what?

Cheers, Wayne