I was thinking about the Tesla PowerWall a lot today, and a lot of that has been in perspective of my own solar project that many of you have been keeping an eye on. So, prepare for a bit of a ramble...
I think there is a large distinction that isn't readily made clear by the presentation and the current website for the PowerWall.
The PowerWall page currently says 10kWh for $3500, or $350/kWh. Sounds pretty decent considering it has integrated BMS, cooling, etc, right? However, the caveat makes this is a mostly useless product for basically everything last night's presentation was about.
Let's focus on that for a moment. So it's basically a 2kW standby generator with 5 hours of fuel at the rate (10 hours at 1kW, 20 at 1/2 kW, etc). Well, that's probably good if you have a very small mandatory loads panel for your refrigerator and a few lights... but not much, if anything else. It's also only good if you get a lot of power outages and they generally last for short periods. Be without power for more than a day, tops, and a PowerWall will be a PowerlessWall. You could combine this with a half decent solar setup to extend this, but people looking at it for backup purposes (perhaps thinking they could replace a standby generator) are going to be pretty disappointed when their electrician tells then they'll need five or more of them to come close to what a standby genset will handle. Keep in mind, these units only put out 2kW. You can get a 2kW standalone gasoline generator for something like $200 nowadays...
So for backup power I think it has severely limited applications as a single unit. It would take several to equal a small standby generator and people without solar would have no way to top off the tank, so to speak, and would be dead in the water after the packs were depleted.
Now I know, Elon means for this to be used with renewables. Makes sense. This would be fine for someone who already has or is getting solar if they want to be able to run their refrigerator and a few lights while the grid is down. I think the backup version, however, has some pretty narrow use cases in the current market.
So let's look at the real product. The unit that can be cycled daily for balancing your power use and shifting power use to off-peak and all of the other wonderful things that were mentioned in the presentation that have absolutely nothing to do with the $3500 unit.
The 7kWh unit is the one that can be cycled daily and be used for these things, and costs $3000 or about $429/kWh. $429/kWh is pretty expensive. To get that money back at average USA utility rates of $0.1229 per kWh you'd need to use that $429 kWh about 3500 times just break even on the unit cost. At a daily cycle that means load shifting 7kWh per day at 100% efficiency (site claims 92%) for almost 10 years just to be neutral with the cost of the 7kWh unit. But wait, there's more...
The PowerWall site specifically says that the unit does not include an inverter. (I ignored this for the previous backup use cases above because it's impractical even without adding this on top.)
If you have grid tied solar already you may already have a grid-interactive inverter that is capable of accepting power from the PowerWall and feeding it into the grid as you desire. Great. It'll still take you over 10 years, average, just to cover the unit cost.
For those who do not, or are thinking about just shifting loads from on-peak to off-peak, you'll need an inverter. A "dumb" 2kW grid-interactive inverter isn't all that expensive, but it will lengthen your ROI time none the less. Get one that can actually be smart about how it handles this battery power and you could approach doubling that time.
This is all assuming the PowerWall unit has within itself the means to charge from the grid. If you need an AC->DC converter, then add in that cost (unknown at this time if the PowerWall has this).
Now, let's say you've gone through all of this and you're still ready to jump on board.
The EIA says that the average American residence consumes about 10,908 kWh worth of electricity per year. That comes out to about 30 kWh per day, or an average load of about 1.25kW 24 hours per day, 7 days per week. The load will obviously vary, but we'll use this as a base. Using this, we can determine that it is likely that during the on-peak, off-peak, and solar peak times of day that there will be enough time in any of them to charge or discharge this 7kWh battery pack fully in any of those three time frames for the average American.
Hypothetical: Let's say that the household has a large enough solar array to produce their yearly kWh usage over the course of a year so their net usage is 0 kWh. Nice! Well, now the utility company gets smart (like many are) and isn't paying this household the same per kWh feed in as they are charging for feed out. We'll say they're nice and only charging $0.10 more per kWh used than sold.
Well this household is in a good position to benefit from the Tesla PowerWall, right? Let's run some numbers.
Assuming a good solar day, the Tesla PowerWall would be fully discharged at the beginning of the day because night time usage would have exceeded 7kWh on average. During the day enough solar power is generated so that all household loads are powered by local generation, and an 7kWh of what would normally have been put into the grid would be generated to charge the Tesla PowerWall. For the sake of simplicity we'll ignore what may have been generated beyond that and just say it was self consumed (best case).
The sun goes down and it's go time. Good stuff. Tesla PowerWall uses it's power to offset loads........ up to 2kW. Anything beyond that will still need to buy power from the grid (HVAC, cooking, hot water, EV charging, etc). But, we'll still be able to offset 7kWh worth regardless over the next 3.5+ hours. Now, 7kWh of usage later we are back to being dependent on the grid. No problem... we just saved $0.70 today, right?
$0.70/day payback on $3000 + inverter + install................. not looking so good for ROI at 12+ years.
Now these are all just contrived examples. What would it really take for this to make sense?
To ROI in 7 years (a reasonable long term investment, beyond this it's best to invest elsewhere usually) the unit would have to save about $1.18/day @ $3000. This isn't counting installation and other equipment needed. At a daily cycle, 7kWh per day, those kWh would each need to save at least $0.168 after efficiency losses. This would need to happen every single day, 100% discharge, 100% charge, to payback in 7 years. And this is just to get your money back.
So, if you have solar and you get paid at least $0.17 less per kWh fed into the grid as you pay per kWh out of the grid, this could work for you. Unfortunately, this is a small subset of solar users so far. A lot of incentives work the opposite and pay more for solar input than charge output. Most just credit the same both ways (net metering) currently. If you're on time of use metering and your on peak is $0.17 more than off peak, and you already have the rest of the equipment needed to make this work for arbitrage, then this could work for you.
The above, in my research, appear to be a pretty small subset of residences. I'm not sure how this can appeal, financially, to users who don't already have the related and required equipment to make this useful.
"But wk, what about using these to go off grid like Elon said?"
First, let's get one thing straight. One of these units is not going to let any normal user go completely off grid. Not one person. Don't kid yourself. 7kWh per day is 23% of what the average American household uses. Double that for daytime solar generation and consumption and you're still at less than half of average. Sure, some people can make that lifestyle work, but that's not mass market. I'm not talking about the people who are willing to make large changes in usage to fit this product into their lives. I'm talking about the regular people who just want something that works for them, not the other way around.
Then, you have to factor that there are days where there is little to no sun. (My worst solar day since testing has been 4% of expected average output... which is really bad and might as well be nothing.) You're going to need a couple of days of reserve power. At half of the average American usage (15kWh) and two days of reserve (probably bare minimum for an off-grid use case) you're looking at 4 or 5 of these PowerWall units *minimum* at half of average usage. Moving back to the average USA residence and that's more like the maximum of nine units just to have normal use + a "two rainy days in a row" reserve. Oh, and nine of these is $27,000 and is almost as powerful as 100A grid service (usually the feed for small homes with non-electric HVAC and cooking/hot water).
Long story short, without major energy efficient life style changes most people would not be able to go 100% off grid even with nine PowerWalls. You could get really close, but at some point you'll run out.
I have avoided comparisons to my own large off-grid solar project to keep this write up as general as possible. However, I feel that it's worth pointing out that my complete setup including everything (100+ solar panels, 8 inverters, installation, and 191kWh worth of Tesla batteries) is costing somewhere around $500/kWh total per kWh of energy storage. Extracting just the parts that the $3000 PowerWall does for that price, those parts cost about $250/kWh (mainly the batteries). Even if my setup were lead acid based (the "ugly" and "smelly" batteries that Elon compared to in his presentation), which it doesn't, the cost would still be significantly less per kWh than a complete system using PowerWalls.
Not only that, many solar installers already do these battery setups. They may not be in a sleek and fancy case like Tesla's, but your wallet will be better for it.
I have 191 kWh of storage (that's the same as ~28 of Tesla's 7kWh units......) because that's what it will take to get me through a string of a few rainy days at average or slightly below average usage and to stay off grid as long as possible. Grid electricity is only $0.10 per kWh where I am, so no amount of grid arbitrage would ever pay for itself here. Honestly, my project is beyond what I probably should have invested in it. Unfortunately I was already too far into it to not follow through once that became apparent.
I'm still confused as to why the unit *doesn't* include a grid interactive inverter (similar to a small Outback Radian)... if it did, at the same price point, I may be writing something entirely different.
Anyway, the TLDR summary version: PowerWall sounds cool, it just doesn't really seem practical for the masses. It makes no sense for the average person to put one of these daily cycle units on their garage wall, especially considering the cost of the additional equipment needed to make it work as advertised. It sounds good at first glance, but if you really run the numbers for your situation.... it probably doesn't make as much sense as you might think.
Feel free to post some numbers for your situation (on-peak/off-peak rates and times, solar buy sell prices/etc) and I'll do some crunching for you if you like in the thread and we'll see how far off I am with all of this.
I hope my dissenting opinion on the PowerWall doesn't bring about immediate hatred. I think that most people who have followed my projects here and elsewhere would know that at least somewhat know what I'm talking about and I wouldn't take such an opinion blindly.
(Keep in mind this is about the PowerWall residential unit. I do believe the commercial unit (PowerPack) has huge potential.)
----
Update: Posted first round of responses to questions and comments here: Elon, I love you... but the PowerWall isn't that great...... yet. - Page 3
Update: Did up a spreadsheet demonstrating an on peak/off peak load shift using two powerwalls and 1000kWh per month normal usage (and example provided by another user). Two PowerWalls, 14kWh/day load shift to off-peak - Google Sheets
(scroll right to see realistic with efficiency calcs)
I think there is a large distinction that isn't readily made clear by the presentation and the current website for the PowerWall.
The PowerWall page currently says 10kWh for $3500, or $350/kWh. Sounds pretty decent considering it has integrated BMS, cooling, etc, right? However, the caveat makes this is a mostly useless product for basically everything last night's presentation was about.
- Caveat: The 10kWh unit is "For backup applications."
Let's focus on that for a moment. So it's basically a 2kW standby generator with 5 hours of fuel at the rate (10 hours at 1kW, 20 at 1/2 kW, etc). Well, that's probably good if you have a very small mandatory loads panel for your refrigerator and a few lights... but not much, if anything else. It's also only good if you get a lot of power outages and they generally last for short periods. Be without power for more than a day, tops, and a PowerWall will be a PowerlessWall. You could combine this with a half decent solar setup to extend this, but people looking at it for backup purposes (perhaps thinking they could replace a standby generator) are going to be pretty disappointed when their electrician tells then they'll need five or more of them to come close to what a standby genset will handle. Keep in mind, these units only put out 2kW. You can get a 2kW standalone gasoline generator for something like $200 nowadays...
So for backup power I think it has severely limited applications as a single unit. It would take several to equal a small standby generator and people without solar would have no way to top off the tank, so to speak, and would be dead in the water after the packs were depleted.
Now I know, Elon means for this to be used with renewables. Makes sense. This would be fine for someone who already has or is getting solar if they want to be able to run their refrigerator and a few lights while the grid is down. I think the backup version, however, has some pretty narrow use cases in the current market.
- Daily Cycling 7kWh Unit
So let's look at the real product. The unit that can be cycled daily for balancing your power use and shifting power use to off-peak and all of the other wonderful things that were mentioned in the presentation that have absolutely nothing to do with the $3500 unit.
The 7kWh unit is the one that can be cycled daily and be used for these things, and costs $3000 or about $429/kWh. $429/kWh is pretty expensive. To get that money back at average USA utility rates of $0.1229 per kWh you'd need to use that $429 kWh about 3500 times just break even on the unit cost. At a daily cycle that means load shifting 7kWh per day at 100% efficiency (site claims 92%) for almost 10 years just to be neutral with the cost of the 7kWh unit. But wait, there's more...
The PowerWall site specifically says that the unit does not include an inverter. (I ignored this for the previous backup use cases above because it's impractical even without adding this on top.)
If you have grid tied solar already you may already have a grid-interactive inverter that is capable of accepting power from the PowerWall and feeding it into the grid as you desire. Great. It'll still take you over 10 years, average, just to cover the unit cost.
- Grid Arbitrage
For those who do not, or are thinking about just shifting loads from on-peak to off-peak, you'll need an inverter. A "dumb" 2kW grid-interactive inverter isn't all that expensive, but it will lengthen your ROI time none the less. Get one that can actually be smart about how it handles this battery power and you could approach doubling that time.
This is all assuming the PowerWall unit has within itself the means to charge from the grid. If you need an AC->DC converter, then add in that cost (unknown at this time if the PowerWall has this).
- Real world numbers
Now, let's say you've gone through all of this and you're still ready to jump on board.
The EIA says that the average American residence consumes about 10,908 kWh worth of electricity per year. That comes out to about 30 kWh per day, or an average load of about 1.25kW 24 hours per day, 7 days per week. The load will obviously vary, but we'll use this as a base. Using this, we can determine that it is likely that during the on-peak, off-peak, and solar peak times of day that there will be enough time in any of them to charge or discharge this 7kWh battery pack fully in any of those three time frames for the average American.
Hypothetical: Let's say that the household has a large enough solar array to produce their yearly kWh usage over the course of a year so their net usage is 0 kWh. Nice! Well, now the utility company gets smart (like many are) and isn't paying this household the same per kWh feed in as they are charging for feed out. We'll say they're nice and only charging $0.10 more per kWh used than sold.
Well this household is in a good position to benefit from the Tesla PowerWall, right? Let's run some numbers.
Assuming a good solar day, the Tesla PowerWall would be fully discharged at the beginning of the day because night time usage would have exceeded 7kWh on average. During the day enough solar power is generated so that all household loads are powered by local generation, and an 7kWh of what would normally have been put into the grid would be generated to charge the Tesla PowerWall. For the sake of simplicity we'll ignore what may have been generated beyond that and just say it was self consumed (best case).
The sun goes down and it's go time. Good stuff. Tesla PowerWall uses it's power to offset loads........ up to 2kW. Anything beyond that will still need to buy power from the grid (HVAC, cooking, hot water, EV charging, etc). But, we'll still be able to offset 7kWh worth regardless over the next 3.5+ hours. Now, 7kWh of usage later we are back to being dependent on the grid. No problem... we just saved $0.70 today, right?
$0.70/day payback on $3000 + inverter + install................. not looking so good for ROI at 12+ years.
- What would it take to be useful?
Now these are all just contrived examples. What would it really take for this to make sense?
To ROI in 7 years (a reasonable long term investment, beyond this it's best to invest elsewhere usually) the unit would have to save about $1.18/day @ $3000. This isn't counting installation and other equipment needed. At a daily cycle, 7kWh per day, those kWh would each need to save at least $0.168 after efficiency losses. This would need to happen every single day, 100% discharge, 100% charge, to payback in 7 years. And this is just to get your money back.
So, if you have solar and you get paid at least $0.17 less per kWh fed into the grid as you pay per kWh out of the grid, this could work for you. Unfortunately, this is a small subset of solar users so far. A lot of incentives work the opposite and pay more for solar input than charge output. Most just credit the same both ways (net metering) currently. If you're on time of use metering and your on peak is $0.17 more than off peak, and you already have the rest of the equipment needed to make this work for arbitrage, then this could work for you.
The above, in my research, appear to be a pretty small subset of residences. I'm not sure how this can appeal, financially, to users who don't already have the related and required equipment to make this useful.
- Going off-grid with PowerWall
"But wk, what about using these to go off grid like Elon said?"
First, let's get one thing straight. One of these units is not going to let any normal user go completely off grid. Not one person. Don't kid yourself. 7kWh per day is 23% of what the average American household uses. Double that for daytime solar generation and consumption and you're still at less than half of average. Sure, some people can make that lifestyle work, but that's not mass market. I'm not talking about the people who are willing to make large changes in usage to fit this product into their lives. I'm talking about the regular people who just want something that works for them, not the other way around.
Then, you have to factor that there are days where there is little to no sun. (My worst solar day since testing has been 4% of expected average output... which is really bad and might as well be nothing.) You're going to need a couple of days of reserve power. At half of the average American usage (15kWh) and two days of reserve (probably bare minimum for an off-grid use case) you're looking at 4 or 5 of these PowerWall units *minimum* at half of average usage. Moving back to the average USA residence and that's more like the maximum of nine units just to have normal use + a "two rainy days in a row" reserve. Oh, and nine of these is $27,000 and is almost as powerful as 100A grid service (usually the feed for small homes with non-electric HVAC and cooking/hot water).
Long story short, without major energy efficient life style changes most people would not be able to go 100% off grid even with nine PowerWalls. You could get really close, but at some point you'll run out.
- My setup
I have avoided comparisons to my own large off-grid solar project to keep this write up as general as possible. However, I feel that it's worth pointing out that my complete setup including everything (100+ solar panels, 8 inverters, installation, and 191kWh worth of Tesla batteries) is costing somewhere around $500/kWh total per kWh of energy storage. Extracting just the parts that the $3000 PowerWall does for that price, those parts cost about $250/kWh (mainly the batteries). Even if my setup were lead acid based (the "ugly" and "smelly" batteries that Elon compared to in his presentation), which it doesn't, the cost would still be significantly less per kWh than a complete system using PowerWalls.
Not only that, many solar installers already do these battery setups. They may not be in a sleek and fancy case like Tesla's, but your wallet will be better for it.
I have 191 kWh of storage (that's the same as ~28 of Tesla's 7kWh units......) because that's what it will take to get me through a string of a few rainy days at average or slightly below average usage and to stay off grid as long as possible. Grid electricity is only $0.10 per kWh where I am, so no amount of grid arbitrage would ever pay for itself here. Honestly, my project is beyond what I probably should have invested in it. Unfortunately I was already too far into it to not follow through once that became apparent.
- Summary, TLDR version
I'm still confused as to why the unit *doesn't* include a grid interactive inverter (similar to a small Outback Radian)... if it did, at the same price point, I may be writing something entirely different.
Anyway, the TLDR summary version: PowerWall sounds cool, it just doesn't really seem practical for the masses. It makes no sense for the average person to put one of these daily cycle units on their garage wall, especially considering the cost of the additional equipment needed to make it work as advertised. It sounds good at first glance, but if you really run the numbers for your situation.... it probably doesn't make as much sense as you might think.
Feel free to post some numbers for your situation (on-peak/off-peak rates and times, solar buy sell prices/etc) and I'll do some crunching for you if you like in the thread and we'll see how far off I am with all of this.
I hope my dissenting opinion on the PowerWall doesn't bring about immediate hatred. I think that most people who have followed my projects here and elsewhere would know that at least somewhat know what I'm talking about and I wouldn't take such an opinion blindly.
(Keep in mind this is about the PowerWall residential unit. I do believe the commercial unit (PowerPack) has huge potential.)
----
Update: Posted first round of responses to questions and comments here: Elon, I love you... but the PowerWall isn't that great...... yet. - Page 3
Update: Did up a spreadsheet demonstrating an on peak/off peak load shift using two powerwalls and 1000kWh per month normal usage (and example provided by another user). Two PowerWalls, 14kWh/day load shift to off-peak - Google Sheets
(scroll right to see realistic with efficiency calcs)
Last edited: