Powerwall 2 or 3?

[justsam]

My daughter is having a new home built in Healdsburg Ca. and asked me to sit in on an upcoming Back up power discussion with the GC and potential providers. There is a 5k square foot house, a 1200 Square foot ADU and a large barn/garage. It is well water and septic. Power is PGE with underground feed and transformer on site, single phase, 240VAC, two 200Amp service. There is no natural gas but there is a large underground propane tank. There is a small vineyard on the property, and a pool/spa with pump equipment.

I know there has been discussion on generator versus battery with the GC but he is not an expert on the subject, but I think due to fewer maintenance and some noise issues, that battery is their preference. Cost is certainly a factor, but she is also of the mind to do it once, do it right, and be done with it. Retro fit with conduit hanging on exterior walls will not work!

To my knowledge there has been one 48 hour outage due to fire risk, and one or two 3 hour outages from PGE. It is an area where PGE has Planned Power Outages to mitigate fire risk. This is in the last 2 year construction phase.

One potential provider has indicated he can do the job with three PowerWalls, however he is also saying he would do it with PW2 as opposed to PW3 because PowerWall 3 is not available right now. So my question is it worth a wait, makes no difference or? Most of the solar panels are already in place and I do not know the provider. What are the right questions to ask during this meeting? One of their vehicles is a Tesla and would be great if excess solar power could be used to charge the car if there is surplus. Also I think a CyberTruck is in their future, and would be great to have a system that can use it's bi-directional charging capabilities and 100kWh battery, since a Tesla SuperCharger with 250kW chargers is 5 miles from their property,

Thanks for any pointers, as I understand basic electrical but have not kept up with solar systems.

 

[infinkc]

I honestly would do it with a Propane Generator(Generac)

Power walls are still so new, and like you see its already going to a 3. The tech is still so new, its still evolving. If they are getting a Cybertruck, i would hold off on the PW and just use that with a generator. The only benefit of the PW would be to store the charge from solar to charge the cars to offset that cost, but the cost of a PW will probably pay for the life of charging, so not even worth it. You have to remember a PW is a battery, just like the cars its only a 10 year life expectancy and with a high price tag. They would need a minimum of 3 PW to to keep the house up and running for that long, thats using bare minimum also.

Generators are reliable and will keep power on as long as its fueled. It will eat up propane if ran for a long period of time, so adding a second tank maybe something they also consider. Would make sure that its only wired to important circuits that they need ON during an outage. No reason to power everything unless needed.

 

[gsmith22]

powerwall 2 and powerwall 3 are different beasts. pw2 is just a batter and does not have an inverter built-in whereas pw3 does. accordingly, pw2 could be used in a retorfit scenario where panels/inverters are already in place and a battery is being added. pw3 is really meant to be used in a new build where the inverter for the panels is built into the pw3. becasue of these differences, its my understanding that pw2 and pw3 will both continued to be offered for the forseable future - pw3 isn't giong to phase out the pw2 because they are different things meant for different situations. sounds like you don't have the right people involved if they don't understand this basic premise.

 

[loganb]

As it's CA and new build, solar is likely required so with that I think onsite storage via battery bank of whatever make you want is required to get the benefit out of the solar. Battery with generator is of course an option, but with onsite solar with storage capacity, I'd probably be less concerned with adding generator. Future proofing by at least putting the conduit and an open box for where a future generator plug could go to make plugging in and feeding the panels easy.

 

[dcg9381]

Here's a comparison. Capacity is about the same, but you can string more than 3 of V2 versions together.

powerwall 2 and powerwall 3 are different beasts. pw2 is just a batter and does not have an inverter built-in whereas pw3 does.

Tesla says the same thing, Google AI lied to me:

I know there has been discussion on generator versus battery with the GC but he is not an expert on the subject, but I think due to fewer maintenance and some noise issues, that battery is their preference. Cost is certainly a factor, but she is also of the mind to do it once, do it right, and be done with it. Retro fit with conduit hanging on exterior walls will not work!

For the cost of 3 power wall 2's and an inverter, you could probably buy a generator and 10+ years of "hands off service". But yea, propane generators are loud. If CA requires this stuff anyway, that shifts the economics though.

To my knowledge there has been one 48 hour outage due to fire risk, and one or two 3 hour outages from PGE. It is an area where PGE has Planned Power Outages to mitigate fire risk. This is in the last 2 year construction phase.

The PW will do better than the generator in terms of "switch speed" (might not even notice an outage when it happens). The problem is power capacity.

3 x 13kWh = 39 kWh. I looks at my power use yesterday and it's currently mild climate here. I pulled 37.3 kWh. So you'll get 22-24 hours out of that thing, give or take if you draw power like me.

 

[bwringer]

Power walls are still so new...

They've been around since 2015, and last I checked it's late 2024 right now.

As it's CA and new build, solar is likely required so with that I think onsite storage via battery bank of whatever make you want is required to get the benefit out of the solar. Battery with generator is of course an option, but with onsite solar with storage capacity, I'd probably be less concerned with adding generator. Future proofing by at least putting the conduit and an open box for where a future generator plug could go to make plugging in and feeding the panels easy.

Yah, adding the "stubout" (or whatever sparkies call it) for future possible generators, battery storage, and/or solar capacity would be very smart.

Tesla says the same thing, Google AI lied to me:

WHAAAAAA? AI is full of ****? When did that start?

Seriously... on any subject I know reasonably well, I've not yet seen one correct answer from that Google AI search garbage. I have no doubt people are getting killed and going bankrupt every day by believing any of this machine-generated BS. Garbage in, garbage out is as true now as it ever was, even though their AI is getting dangerously good at generating believable sentences. It's goddamn irresponsible of Google to put that **** at the top of every search result, with no way to flag it for in accuracy or dangerous misinformation.

But I digress.

The PW will do better than the generator in terms of "switch speed" (might not even notice an outage when it happens). The problem is power capacity.

3 x 13kWh = 39 kWh. I looks at my power use yesterday and it's currently mild climate here. I pulled 37.3 kWh. So you'll get 22-24 hours out of that thing, give or take if you draw power like me.

You do need something to tell you you're drawing on your PW, and to automatically or manually drop loads that aren't essential. You can stretch that amount of capacity a long way if you can hold off on doing laundry, turn off the A/C, stove, and lights you don't need, maybe finish the movie later, etc.

 

[dcg9381]

You do need something to tell you you're drawing on your PW, and to automatically or manually drop loads that aren't essential. You can stretch that amount of capacity a long way if you can hold off on doing laundry, turn off the A/C, stove, and lights you don't need, maybe finish the movie later, etc.

Yes, you'd need a notification that power is down because these things switch so fast (sub second) that you'll likely not know that the power is down. And your notification is probably "blown" if the local internet is offline.

The other thing I catch is that the PW2 is only good for 5kW output continuous. So 3 of 'em is kinda a good "minimum to get started number". If you've only got 15kW available, I assume that part of the install is to design load shed or NOT protect all circuits because it's not to cover the whole panel.

 

[justsam]

Thanks to all for your input. In spite of this being a green field build I sense that expectations were not well understood. The intent of the system was of course to maximize solar during the day, and make it the first choice energy source. Batteries to come into play during low solar output, and PGE as a last resort. All of this to offset what was estimated to be a $1200/month PGE bill. The solar system was sized beyond the CA. minimum in order to accommodate this. The other expectation is to have back up power and that appears to be the issue if only batteries are used. New meetings are being scheduled with all the existing players and perhaps some new ones as directed by home owner and GC. No one seems ready to discuss EV battery bi-directional back up yet.

 

[ericm]

Our part of California has a lot of power outages and usually 1-3 week long outages a year. Many people in the area have solar and batteries. The system I know that's best designed has a generator input to charge the batteries when there's a long outage in the winter. They heat with propane and top up the batteries every couple days with the generator. The people with just batteries wind up having to run just the fridge and a few lights after a day or two (or three for the guy with $40k worth of batteries).

At every extended outage the local mailing list is full of people with Powerwalls that stopped working. They report how they're doing with Tesla support (usually not well), trade tips on undocumented ways to reboot the batteries that may or may not work, and commiserate about how they had to leave their house and find a room in town. The folks with Enphase don't seem to have problems.

 

[dcg9381]

Thanks to all for your input. In spite of this being a green field build I sense that expectations were not well understood. The intent of the system was of course to maximize solar during the day, and make it the first choice energy source. Batteries to come into play during low solar output, and PGE as a last resort. All of this to offset what was estimated to be a $1200/month PGE bill. The solar system was sized beyond the CA. minimum in order to accommodate this.

I can speak to this a bit. Note, I'm not a Tesla installer. It's my understanding that the Power Wall (or likely the right inverter with a PW2) can be "programmed" to draw down battery power instead of using from the grid. If you're on "variable" power costs, you set it to charge during "low cost times" and discharge at "high cost times".

In situations (like mine and I suspect most of CA) the "payback" of feeding the grid is a fraction of what you pay for power delivery per kWh, so the "only" way to make that math work is with batteries and programming to use PV And "charge low" / discharge "high" (cost timing)... This implementation, the programming is going to be very important to savings.

How big is PV the array? Array power is "not stable" (all it takes is a single cloud) - so you buffer through a battery. 15kW of power throughput isn't enough to cover a 200A panel... Not even close.

Here's the other deal, if you're using this installation correctly to optimize your power bill: Drawing the batteries down during the day setting it to charge primarily from PV... It's not going to be in a "ready condition" (necessarily) to handle an extended outage as it's almost always at partial charge state.

Let me know if this makes sense. Proper (programming) configuration in this installation will make a huge difference in long term cost savings.

 

[ericm]

In situations (like mine and I suspect most of CA) the "payback" of feeding the grid is a fraction of what you pay for power delivery per kWh, so the "only" way to make that math work is with batteries and programming to use PV And "charge low" / discharge "high" (cost timing)... This implementation, the programming is going to be very important to savings.

Yes. CA does not have net metering and under the current NEM3 rules the payment for power sent to the grid is tiny. The only way to make solar pay is to use batteries. I'd expect the GC and solar companies know this.

 

[dcg9381]

I'd expect the GC and solar companies know this.

You have high expectations. Solar is very "niche" and I don't know a single GC (including one that's a master sparkie) that has a clue on solar. They just throw it to a sub (like me).

Solar companies... Meh. Yes, they should know in CA the only payback is via battery. They'll get the stuff up and functional no doubt. But it's like the plumber installing my tankless water heater. He installed it, got it to work, but all the magic for configuring pre-heat loops on-demand pre-heat, he didn't touch. Most consumers would never know the difference.

Definitely ASK the solar company about how they plan to accomplish your goals in your specific area. A good one will have cost projections that involve charging batteries at night when cost per kWh is low. And they should talk to you about how they plan to set the inverter to discharge during high kWh times. They should also be telling you that this affects your "backup" capacity. There is a bunch of math here (that can be done with the right software), but the fact that you only have 15kW of capacity - I don't see how that covers the whole panel. I don't know Tesla, so I could be missing something.

I've never built in CA, but I've built several times... Assuming nothing when it comes to building homes.

 

[pembol]

First, I would look beyond Tesla for your solar and battery solution - their customer service is absolutely abysmal. There are many other players in the solar/battery field.

It sounds like your daughter is sizing her solar to run their house, in which case a good design criteria would be to size the battery system to run typical loads for 1 night, and critical loads for 2-3 nights. That way they can be essentially self sufficient of the grid - the solar runs the house and charges the batteries during the day, and then runs off the house at night. If their were a power outage, the system would be able to tide critical loads over several days in case of cloudy days or snow on the panels.

If they have time of use pricing and net metering, you can also use the battery for some arbitrage - preferentially charge the batteries when ever there is solar available (as opposed to preferentially running the house) then sell the power back to the grid during peak periods (typically late afternoon/evening).

We have an all electric house with a 10kW Solar Edge system - it is great, our total utility bill is $7/month. We don't have batteries at the moment as our grid is reliable, but may consider it in the future.

 

[gsmith22]

No one seems ready to discuss EV battery bi-directional back up yet.

the issue isn't with tesla, enphase, franklin, etc. its with the car companies. the cars need equipment to make this work, vey few have it, and the instantanteous power you can draw isn't as good as the powerwalls, enphase, franklin, etc. batteris. i think its like the f150 lighting, nissan leaf, and maybe one or two other cars where you could actually draw from the car's battery in a no power situation. if you need backup power, the home batteries are where it is at. could change in the future but for reasons that aren't clear to me, no one seems interested

 

[dcg9381]

if you need backup power, the home batteries are where it is at. could change in the future but for reasons that aren't clear to me, no one seems interested

You see fewer of these installs because they're a lot more labor and involve 2-3x the hardware costs. Straight PV solar, you can drop in a breaker for back-feed and pretty much be done with the panel work. This is likely going to involve sub-panels.. Which is not a big deal on a new build, but on retrofit, it's expensive.

If you're using this system as a money saving device and leveraging time of use rates, it's not exactly sitting around at full capacity. Want to do 48 hours power back up, that's a LOT of batteries for most homes.

 

[gsmith22]

You see fewer of these installs because they're a lot more labor and involve 2-3x the hardware costs. Straight PV solar, you can drop in a breaker for back-feed and pretty much be done with the panel work. This is likely going to involve sub-panels.. Which is not a big deal on a new build, but on retrofit, it's expensive.

If you're using this system as a money saving device and leveraging time of use rates, it's not exactly sitting around at full capacity. Want to do 48 hours power back up, that's a LOT of batteries for most homes.

I was referring to the use of an EV for backup power. No one seems interested it that relative to just using g standalone home batteries. Otherwise, yes grid tied solar is easy. Paperwork with poco is probably biggest hurdle.