Energy in 2 powerwall drained from 100% down to 5%

[m3-pw2]

I did a test last week to see how much energy can be consumed from 2 powerwalls in case of power outage. My battery was at 100% (due to grid services event), and by the end of the day it was drained down to 5% (0% reserve in app yields 5% in portal). When the reserve reached 5% in portal, the powerwall stopped powering the house. I assume this might be the same behavior when there is power outage.

For those trying to figure out how much actual energy each powerwall can provide them during a power outage, here are the numbers.

2x Powerwall2:
- Energy provided from 100% -> 5%: 22.2 kWh (so 11.1 kWh per Powerwall)
- Energy consumed to recharge 4% -> 99%: 24.6 kWh
- Round-trip loss: 9.76%

What is interesting is that I am not getting the advertised 13.5 kWh usable battery energy. In theory, 95% energy with 5% single-trip loss should yield 12.18 kWh per Powerwall. So I am short of 1 kWh. I have been told that the Powerwall keeps 1 kWh of storage to avoid fully draining the battery which would require a technician visit to bring it back to life. Can anyone else confirm this?

 

[bkp_duke]

No one ever gets close to the usable number because of the safeguards in place to keep the system from damaging the batteries.

Your numbers (discharge kwh, recharge kwh, and round trip efficiency) are spot on and are what others in this forum have observed.

 

[Keith]

I did a test last week to see how much energy can be consumed from 2 powerwalls in case of power outage. My battery was at 100% (due to grid services event), and by the end of the day it was drained down to 5% (0% reserve in app yields 5% in portal). When the reserve reached 5% in portal, the powerwall stopped powering the house. I assume this might be the same behavior when there is power outage.

For those trying to figure out how much actual energy each powerwall can provide them during a power outage, here are the numbers.

2x Powerwall2:
- Energy provided from 100% -> 5%: 22.2 kWh (so 11.1 kWh per Powerwall)
- Energy consumed to recharge 4% -> 99%: 24.6 kWh
- Round-trip loss: 9.76%

What is interesting is that I am not getting the advertised 13.5 kWh usable battery energy. In theory, 95% energy with 5% single-trip loss should yield 12.18 kWh per Powerwall. So I am short of 1 kWh. I have been told that the Powerwall keeps 1 kWh of storage to avoid fully draining the battery which would require a technician visit to bring it back to life. Can anyone else confirm this?

Don’t forget inverter losses. It’s why it’s always better to use power directly from solar vs charging the Powerwalls and then using the energy later.

 

[gpez]

No one ever gets close to the usable number because of the safeguards in place to keep the system from damaging the batteries.

This is not correct. "Usable" by definition accommodates the battery protection limits.

A brand new Powerwall 2 has a 14kWh total energy rating and 13.5kWh usable as per the Powerwall 2 data sheet. @m3-pw2 since you have 2x Powerwalls this could be the 1kWh reserve you're thinking of.

@m3-pw2 you can also check what your Powerwalls are reporting as their maximum capacity in wh through the Tesla cloud APIs. As the Powerwall ages the capacity diminishes and it is warrantied to retain 70% capacity over 10 years depending on operating mode as per the official warranty document.

Details on how to query the Tesla Cloud APIs to see what your Powerwalls are reporting their maximum capacity as is here: Any way to monitor battery degradation?

 

[bkp_duke]

This is not correct. "Usable" by definition accommodates the battery protection limits.

A brand new Powerwall 2 has a 14kWh total energy rating and 13.5kWh usable as per the Powerwall 2 data sheet. @m3-pw2 since you have 2x Powerwalls this could be the 1kWh reserve you're thinking of.

@m3-pw2 you can also check what your Powerwalls are reporting as their maximum capacity in wh through the Tesla cloud APIs. As the Powerwall ages the capacity diminishes and it is warrantied to retain 70% capacity over 10 years depending on operating mode as per the official warranty document.

Details on how to query the Tesla Cloud APIs to see what your Powerwalls are reporting their maximum capacity as is here: Any way to monitor battery degradation?

I don't care what the definition is, I'm telling you the REAL WORLD observed results.

We can argue about them all day long, but that is the simple fact of the matter.

And I've already done the degradation calculation, it's only 3-5% on my powerwalls.

 

[gpez]

I don't care what the definition is, I'm telling you the REAL WORLD observed results.

We can argue about them all day long, but that is the simple fact of the matter.

And I've already done the degradation calculation, it's only 3-5% on my powerwalls.

Easy there. I wasn't stating that you aren't getting less than the 13.5kWh fact sheet usable - I get less than the 13.5kWh fact sheet usable. The last API call I made showed about 12.2kWh. I have an open ticket with Tesla to explain to me why. My money is on them coming back and saying "that's normal degradation after 1 year".

What I am stating is that your "safeguards" explanation for why your observed usable is different than the published usable is incorrect as the definition of usable accounts for that.

If you're getting substantially less usable out than what the degradation would account for then you need to get a real explaination from Tesla as to why rather than asserting things that contradict Tesla's own published literature.

I'm also curious how you've already done the degradation calculation and only came up with 3-5% if you're asserting there's some hidden safeguard reserve not accounted in the usable figure.

 

[JayClark]

I'd also note, that the PWs just being "on" 24 hours a day, in my case with 4 powerwalls, my PWs will consume about 2k per day even if never charged or discharged. I've measured mine to use between 20-25 watts each on average (they seem to periodically gulp some juice for maintenance purposes during non-use periods). Given you have 2 PWs and you are coming up short about 1k.... that's probably just the remaining operating/maintenance/communication usage for the day, for the two powerwalls, separate from any other charge/charge-down activities. Or, at least that would match what I see for my four Powerwalls when I've measured carefully directly with a whole house meter and compared to what the tesla app is reporting on days when I'm not charging/discharging the PWs like on the weekends.

 

[SnakesOnAPlane]

I have 4 Powerwalls and have my cutout at 20%. In hot summer days (Florida), as I run the A/C at 68-70 deg from 9pm to 7am, I may hit 20% before the sun comes up. Often in the summer, we have a bunch of cloudy days and rain, so I’m not always recharging the batteries to 100% before sundown. It varies...

 

[RobbL]

When you did your test, was there a setting you used to prevent the Powerwalls from recharging from your solar panels or the grid? When our power company here in Northern California turns off our power in the Fall for one or two days due to fire danger, I will be relying on recharging from our solar panels during the daytime to extend the time we can get by.

 

[m3-pw2]

When you did your test, was there a setting you used to prevent the Powerwalls from recharging from your solar panels or the grid? When our power company here in Northern California turns off our power in the Fall for one or two days due to fire danger, I will be relying on recharging from our solar panels during the daytime to extend the time we can get by.

The Powerwall doesn’t charge from grid unless there is a storm watch. I did the test at the end of the day and the batteries did not charge from solar until next day.
We lost power two days ago when tropical storm hit our area. We don’t expect power to be restored for 5-7 days. So now I am learning how they work during outage. It’s nice to be running lights without the noise of generators.

 

[m3-pw2]

I'd also note, that the PWs just being "on" 24 hours a day, in my case with 4 powerwalls, my PWs will consume about 2k per day even if never charged or discharged. I've measured mine to use between 20-25 watts each on average (they seem to periodically gulp some juice for maintenance purposes during non-use periods). Given you have 2 PWs and you are coming up short about 1k.... that's probably just the remaining operating/maintenance/communication usage for the day, for the two powerwalls, separate from any other charge/charge-down activities. Or, at least that would match what I see for my four Powerwalls when I've measured carefully directly with a whole house meter and compared to what the tesla app is reporting on days when I'm not charging/discharging the PWs like on the weekends.

With two batteries, the average daily loss is 0.5kWh per day. For 4 batteries, 2kWh per day seems a lot than expected.
When I did the test, the level was kept at 100% until 4pm. At 4pm, they started offloading power to the grid to help with peak demand and reached 20% at 7pm. At that point, I switched it to self powered and set reserve to 0% in app and it drained all the way down by 9pm. Portal showed 5% remaining at the end.
Next day it started recharging from solar and was done by 12:25pm. It depleted to 4% overnight and only charged up to 99%.
For energy usage, I think that the PW draw from 4pm to 9pm should be insignificant. 0.5kWh*5 hours/25h=0.10kWh.

 

[m3-pw2]

Don’t forget inverter losses. It’s why it’s always better to use power directly from solar vs charging the Powerwalls and then using the energy later.

Inverted losses are already accounted for in the round trip losses I mentioned in the post. I got 10% loss for the full round trip. I assumed it is half when it is one way.

 

[m3-pw2]

This is not correct. "Usable" by definition accommodates the battery protection limits.

A brand new Powerwall 2 has a 14kWh total energy rating and 13.5kWh usable as per the Powerwall 2 data sheet. @m3-pw2 since you have 2x Powerwalls this could be the 1kWh reserve you're thinking of.

@m3-pw2 you can also check what your Powerwalls are reporting as their maximum capacity in wh through the Tesla cloud APIs. As the Powerwall ages the capacity diminishes and it is warrantied to retain 70% capacity over 10 years depending on operating mode as per the official warranty document.

Details on how to query the Tesla Cloud APIs to see what your Powerwalls are reporting their maximum capacity as is here: Any way to monitor battery degradation?

I wasn’t aware that Tesla api exposed capacity and other useful data. I’ll need to figure out how to query it. Thanks for the link.

 

[wjgjr]

With two batteries, the average daily loss is 0.5kWh per day. For 4 batteries, 2kWh per day seems a lot than expected.
When I did the test, the level was kept at 100% until 4pm. At 4pm, they started offloading power to the grid to help with peak demand and reached 20% at 7pm. At that point, I switched it to self powered and set reserve to 0% in app and it drained all the way down by 9pm. Portal showed 5% remaining at the end.
Next day it started recharging from solar and was done by 12:25pm. It depleted to 4% overnight and only charged up to 99%.
For energy usage, I think that the PW draw from 4pm to 9pm should be insignificant. 0.5kWh*5 hours/25h=0.10kWh.

I think the biggest thing that would affect PW loss is their location, particularly temperature. If, for example, they are outside in the hot sun all day, they may need to run their fan a lot, which could account for the additional loss. That said, our losses (2 PWs) are much closer to the 0.5kWh number, and possibly below - I have been logging the data so at some point will have to run the numbers, but that seems to be about what is happening. But our PWs are inside, mounted on a basement wall, so the temperature should be relatively close to ideal.

Interestingly, when I did a test with draining the PWs a couple months back, power to our house was shut off when the API indicated 10% (5% on the app).

 

[bkp_duke]

With two batteries, the average daily loss is 0.5kWh per day. For 4 batteries, 2kWh per day seems a lot than expected.
When I did the test, the level was kept at 100% until 4pm. At 4pm, they started offloading power to the grid to help with peak demand and reached 20% at 7pm. At that point, I switched it to self powered and set reserve to 0% in app and it drained all the way down by 9pm. Portal showed 5% remaining at the end.
Next day it started recharging from solar and was done by 12:25pm. It depleted to 4% overnight and only charged up to 99%.
For energy usage, I think that the PW draw from 4pm to 9pm should be insignificant. 0.5kWh*5 hours/25h=0.10kWh.

Losses also vary by temperature. Batteries in say a garage with a narrow temperature range should have less losses than those exposed to more extreme temperatures outside. Just like Tesla cars, the powerwalls have BMS systems that keeps their temps regulated.

 

[m3-pw2]

Losses also vary by temperature. Batteries in say a garage with a narrow temperature range should have less losses than those exposed to more extreme temperatures outside. Just like Tesla cars, the powerwalls have BMS systems that keeps their temps regulated.

That's a good point @bkp_duke. Mine is in the basement as well.

 

[m3-pw2]

I think the biggest thing that would affect PW loss is their location, particularly temperature. If, for example, they are outside in the hot sun all day, they may need to run their fan a lot, which could account for the additional loss. That said, our losses (2 PWs) are much closer to the 0.5kWh number, and possibly below - I have been logging the data so at some point will have to run the numbers, but that seems to be about what is happening. But our PWs are inside, mounted on a basement wall, so the temperature should be relatively close to ideal.

Interestingly, when I did a test with draining the PWs a couple months back, power to our house was shut off when the API indicated 10% (5% on the app).

@wjgjr My powerwalls are in the basement too. So we get the same daily consumption of around 0.5 kWh.
That is troubling that the PW shut down at 5% in app, or 10% API. I was already not liking those numbers that I am getting. Tesla advertises that these PWs have usable 13.5kWh. It seems more and more that in practice, usable is much less than that. They should have been more transparent about what to expect.

 

[JayClark]

With two batteries, the average daily loss is 0.5kWh per day. For 4 batteries, 2kWh per day seems a lot than expected.
When I did the test, the level was kept at 100% until 4pm. At 4pm, they started offloading power to the grid to help with peak demand and reached 20% at 7pm. At that point, I switched it to self powered and set reserve to 0% in app and it drained all the way down by 9pm. Portal showed 5% remaining at the end.
Next day it started recharging from solar and was done by 12:25pm. It depleted to 4% overnight and only charged up to 99%.
For energy usage, I think that the PW draw from 4pm to 9pm should be insignificant. 0.5kWh*5 hours/25h=0.10kWh.

As other mentioned by others, local temperatures can increase losses as the PW works harder to maintain health and such. I can "hear" the difference when it's hot in my garage, vs under the same load when it's 20 degrees cooler. I live in Arizona, and I know definitely have more losses during "idle" times that what you're seeing, at least in the summer.

But it's roughly what I expected (even if I hoped for slightly better) after having experience with inverters and charger controllers of various types over the years, running in my local conditions. The PWs still are more efficient than any other system I've used, and none of those could power my entire home, ac and all. I also only use the PWs to power my home directly - so there may be some operational nuances when PW are used to feed back to the grid that cause different behavior and performance vs when just powering incremental and variable loads that may on average be lower.

In regard to peak rated efficiency, efficiency of these types of products is usually at specific point in a products total performance curve. Usually that point is targeted to be in the sweet spot of the "average" user and usage. So operating to either side of that sweet spot results in progressively lower efficiency. I'm totally pulling this out of my backside, but for instance, while the PW-2s are rated at being capable of achieving 5kW continuous and 7kW peak, it may be that max peak efficiency (as quoted in specs) is at what Tesla considers ideal conditions and loads (this is true of most products I've used). So maybe the PWs only achieve 95% peak efficiency at the ideal ambient temperature of 80 degrees, with a continuous draw of say 50% of max load, so maybe say when running at 3.5 kwHrs. Not sure if that's the case here, but jives with other systems I've used.

 

[m3-pw2]

As other mentioned by others, local temperatures can increase losses as the PW works harder to maintain health and such. I can "hear" the difference when it's hot in my garage, vs under the same load when it's 20 degrees cooler. I live in Arizona, and I know definitely have more losses during "idle" times that what you're seeing, at least in the summer.

But it's roughly what I expected (even if I hoped for slightly better) after having experience with inverters and charger controllers of various types over the years, running in my local conditions. The PWs still are more efficient than any other system I've used, and none of those could power my entire home, ac and all. I also only use the PWs to power my home directly - so there may be some operational nuances when PW are used to feed back to the grid that cause different behavior and performance vs when just powering incremental and variable loads that may on average be lower.

In regard to peak rated efficiency, efficiency of these types of products is usually at specific point in a products total performance curve. Usually that point is targeted to be in the sweet spot of the "average" user and usage. So operating to either side of that sweet spot results in progressively lower efficiency. I'm totally pulling this out of my backside, but for instance, while the PW-2s are rated at being capable of achieving 5kW continuous and 7kW peak, it may be that max peak efficiency (as quoted in specs) is at what Tesla considers ideal conditions and loads (this is true of most products I've used). So maybe the PWs only achieve 95% peak efficiency at the ideal ambient temperature of 80 degrees, with a continuous draw of say 50% of max load, so maybe say when running at 3.5 kwHrs. Not sure if that's the case here, but jives with other systems I've used.

Fair point regarding inverter efficiency at different loads. During my test, the first 80% were drawn at a constant 6kWh.
I think Tesla should make the inverter efficiency chart available, similar to the chart I have from my solar inverter. At this point I am really unsure how long we could last on 2 PWs without any solar production due to weather.
Since I wrote this post, we had Tropical Storm Isaias in our area and power is out. Damages are more than anticipated, and no one really knows how long it would last. I am able to bring the house consumption to about 0.5kWh. We anticipate rain tomorrow, so we will be extra vigilant in our consumption. But the main question I would like to answer is how long can one go on PWs with no solar production? Assuming I can get the same 22kWh from the PWs, it means I can last 1 day and 20 hours. At least in summer, I haven't seen no solar production at all.

 

[wjgjr]

@wjgjr My powerwalls are in the basement too. So we get the same daily consumption of around 0.5 kWh.
That is troubling that the PW shut down at 5% in app, or 10% API. I was already not liking those numbers that I am getting. Tesla advertises that these PWs have usable 13.5kWh. It seems more and more that in practice, usable is much less than that. They should have been more transparent about what to expect.

I was surprised about the 10% number. I'd test again to see if it was an odd behavior (it was pre-PTO) but I don't think my wife liked the first test (particularly since it went off before I expected.)