Backup Solar Power?

Syonyk suggested that I inquire here about this.

I live in Texas. February was an unpleasant wake up call. We experience outages once every couple years, and what’s really upsetting is being unable to make food because the grid is down. Like many others, I decided to add some form of backup power. I had no idea what I needed, so I got some CT clamps and started measuring things. The short answer is: 1.5kW is the peak draw on my critical circuits. More would be great, because not having AC when it’s 110F is not fun, but it’s infrequent enough that it doesn’t make sense to spend a downpayment on a house when I can run fans for a couple of hours.

My initial thought was to buy a generator, but then I stumbled across the SPS feature on the SMA Sunny Boy. The SB can produce up to 2kW on SPS during an outage. And, unlike a generator, the system pays for itself by producing during the 99.9% of the time that the grid is up. However, the rub here is that SPS only produces single-phase/120V. The loads I want to run are in a subpanel fed with both legs from the mains. Either I would need to short the two legs together, which is a terrible idea, or I would need to produce split-phase/240V from SPS. Victron has a 100A auto-transformer selling for $700 on Amazon, but I have not investigated further.

I am weighing this against a second option, which is to put in an off-grid inverter. I would be able to use the full capacity of the array, but I would need to buy batteries, and batteries are expensive and would need to be replaced. I have an existing grid-tied system, and I was told that the off-grid inverter would allow me to use both systems. This seems surprising, so I don’t know if it’s true.

My question is, is there a better way to do this? I’m concerned with value, not cost, which is why solar seems to me a better option than a generator.

1 Like

Quick thought, for your critical loads, do they include things like fridges/freezers that generate a larger initial startup surge than running? That might blow your power budget. Also don’t forget that just because the SB can output 2kW from SPS, that doesn’t mean it will since it might be night time, or excessively cloudy, or what not. So from that perspective, a battery backed system, even if it’s just a small-ish capacity (as long as it can handle the load) or a generator might be preferable.

Just having read from Syonyk what he’s written so far about those Sunny Boys, and not knowing grid-tie very much, could you run 2 strings of panels and 2 SB to and tie the outputs together to produce each leg to give you the split phase 240v? Or am I totally naive about that and that won’t work?

You say you have an existing solar array?

I do have a fridge, but it’s the magic linear compressor sort. As best I can tell, there is no surge when starting. It usually stairsteps up to 80W, runs for a bit, and then cuts off. It has never exceeded 300W. We mostly cook with Instant Pots, the larger of which pulls 1.1kW. The microwave is about the same. The slow cooker is only 250W, which should be no trouble even on a cloudy day.

This is the first time I’ve ever experienced a multiday outage, and I’m just looking for a way to survive for a few days if it happens again. I’m willing to accept unreliability in exchange for not spending a small fortune. Otherwise, I’d add batteries to my current system and/or buy a generator and done.

I do have a small array running off of a SolarEdge SE5000A.

Powering each leg off of a separate SB is an interesting idea, but the neutrals would need to be bonded, and I’m not sure what that would do. I don’t think it’s necessary, though, as I can run everything within the 2kW SPS budget off of one unit.

Is the entire subpanel critical loads, or are there only a few circuits in the subpanel that matter? If it’s only a few circuits, could you rearrange the panel so all the critical loads were on one phase? Even getting some half height breakers would be cheaper than a lot of other options if you could get everything cleanly onto one phase.

Autotransformers are a thing, but not one I’m terribly familiar with - and I’d be especially curious as to how they handle split loads. If most of your loads are 120V, then you won’t have it balanced, and I’m not sure if there are limits on how unbalanced a load they can still serve. Dig around the datasheets for sure.

As I’ve mentioned elsewhere, you really need to dig into the datasheets and manuals for the proposed off grid inverter solution to understand what they will and won’t do. Handling 10kW of rooftop solar with AC coupling can be done, but it’s not cheap, easy, or likely to get past the plans review process, at least out here… though it would give you the most capability. I’d be surprised if you could get it set up for under $10k, and then you’re into regular battery maintenance for a capability that’s just not going to be used very often. A big propane tank and generator would almost certainly be far cheaper for the capability.

Unfortunately, doing what you want to do is just difficult - I tried and failed with my house system. The original design was for a battery backed, DC coupled, fully off grid capable system that would have allowed me to treat the grid as a “nice to have” - if it’s up, great, I’ll draw from it or export to it, if not, oh well. And the complexities and costs of doing it just started spiraling up beyond “stupid.” I think I quite somewhere around $60k in materials and contracting work (replacing the inside panel for remotely operated breakers, replacing the outside panel, many thousands of dollars of wire to run 200A around because the load sheets say my house needs 200A even though it needs nothing of the sort, etc).

I do not believe that will work - at least, I’ve not seen an easy way of doing it.

One could have a small 48V battery bank, a standalone split phase inverter, and a couple 120V to 48V battery chargers (or 24V, but… eennnh… that’s a lot of amps) to power split phase loads from the SPS outlets, and that’s what I’m planning to do, but that project is currently waiting on batteries, among a few other things that we can’t get right now for various reasons. :confused:

Is the entire subpanel critical loads, or are there only a few circuits in the subpanel that matter? If it’s only a few circuits, could you rearrange the panel so all the critical loads were on one phase? Even getting some half height breakers would be cheaper than a lot of other options if you could get everything cleanly onto one phase.

I’m sure I could, but then I’m doing a lot of extra work rewiring the panel. Most rooms in my house are split randomly across different circuits. It would be quite frustrating to forget which outlets work and which do not. If there were a good reason not to run the whole subpanel off of SPS, then fine, but I can’t think of one. SB asserts that SPS trips off when you try to pull more than 2kW, and I could always throw a breaker in between to be extra safe.

If most of your loads are 120V, then you won’t have it balanced, and I’m not sure if there are limits on how unbalanced a load they can still serve.

As best I could grok the manual, it looks like the 32A version can do 32A on each leg for up to 30 minutes and 28A continuous. SPS is quite a bit less, so unless I’m missing something, this shouldn’t be a problem. But obviously I’d want to confirm that before going that route.

Label the backed up ones.

It will trip off when you try to exceed the rated current (or the available energy from the panels). There’s no need to throw a breaker in the mix - the SPS (and the inverters in general) are a current limited supply, so as long as the wiring is able to handle the maximum output current (12AWG definitely will, 14AWG might - it’s on the edge, though, so check your ampacity charts), a breaker doesn’t add anything. However, depending on how you plan to wire it up, you may need a breaker anyway. But your life will be far easier, permit-and-inspections-wise, if you just hang an outlet on the SPS and a generator inlet plug to the autotransformer.

It does sound like the one you’re looking at would be able to handle the mix of loads, though.

So probably not fitting for your situation, but I wonder if electric car batteries could be used / considered for energy storage. I if you could feed the battery back into your house in the case of outage.

You know, looking into autotransformers, they’re not as expensive as I’d thought. I really should order one and play with it - experiment with that for backfeeding the house loads. We have more than enough panel to run 2kW sustained during any reasonable sun, and I can certainly answer questions about the SPS and autotransformer if I have one! If SPS + autotransformer runs our well pump and scattered other stuff during the day, that’s useful. Even if it takes some manual toggling of breakers to manage load (I’m not sure what the startup surge on the well pump is). On the plus side, big inductor in the form of a transformer.

They can be used for storage, and I’m not aware of a way to do one that meets NEC requirements, which almost certainly means that your insurance company will not cover any incident if they can prove the battery is at all related. I’m aware of the DIY Powerwall folks, think they’re dangerously insane (they habitually abuse lithium and laugh about it), and would love to see how their insurance companies handle structure fires.

You’d need to keep a car battery inside for thermal management (lithium can’t be charged when cold, EVs will burn shore power to heat the battery before charging in the cold weather), and as it’s not going to be a UL or ETL listed energy storage system, and it’s over the “low voltage” exemptions (up to 48V lead acid), good luck getting anyone to sign off on it. It’s pretty explicitly not within the bounds of NEC.

AC coupled storage is a thing, just not a terribly cheap thing. The Tesla system works well enough, if a lot expensive and you can’t install it yourself. Other systems exist as well, but generally have similar issues with “expensive” and “may or may not be able to install it yourself.” I get the impression the OP isn’t looking to drop $20k on a system.

But… hm. Autotransformer, 120V in, a small generator or even an EV based backup system (they usually run 1500W or so off the car’s 12V system)… that really is worth experimenting with.

…I wonder if electric car batteries could be used / considered for energy storage…

This has been proposed at least twice: recycling HEV/PHEV/EV batteries (since grid storage is mostly agnostic toward energy density) and using your EV to create a microgrid. I don’t know why neither is commonplace, but I’m sure there’s either a technical or regulatory reason for it.

Ignoring the regulatory aspect…

Recycling old EV batteries can work, but as @Syonyk said, is pretty dangerous without the right engineering and safety systems.

Ad hoc use of the batteries while they are installed in EVs strikes me as viable, but unwise since most of the value of the batteries/vehicle hinges on the batteries staying high performance and high energy density. I wouldn’t be inclined to add unnecessary drain/cycles for other purposes that could be served by cheaper lower performance batteries/sources.

The proposal was to use the EV’s batteries specifically for backup power in grid-tie systems. If outages are sufficiently rare, then adding the occasional extra discharge cycle to your EV batteries is probably less expensive than buying a proper backup power system. Depending on the requirements, a proper backup system ranges from $500 (small portable generator + extension cord) to >$10,000.

I don’t know what a good number for the cost of EV batteries is, but here’s a wildly pessimistic one: a Tesla Model 3 used as a home storage system using a lifetime of 2,500 cycles. This gives 2,500 * 82 kWh/$40,000 = ~$5.13/kWh/cycle. Even using this wildly overstated figure, you can work out that you come out ahead against an inexpensive portable generator up to the 100~200 kWh mark. The break even point is an order of magnitude higher for a permanently installed generator or even battery-backed hybrid solar inverters.

If you inflate both the useful battery capacity and the useful battery cycles, it doesn’t produce a clear (worst case) “pessimistic result”.

From a little googling, it looks like;

  • 82kWh is the raw battery capacity, so one would expect an upper bound of closer to 72kWh of actual capacity
  • The 82kWh battery only comes in the “long range” version of the Tesla model 3 which has a price tag closer to $50K
  • Tesla only advertises ~1500 cycles

So repeating your math (with the division in the right place);

$50,000 / (1,500 * 72 kWh) = ~$0.46/kWh/cycle

Or a more intuitive way to look at it;

$50,000 / 1,500 = ~$33.33/cycle

So in other words each cycling of the battery could cost ~$30. Thus from a purely economics standpoint, I agree with your premise that incidental backup power usage might make sense.

Personally, I think it’s wiser for most people to just conserve power when they don’t have viable off-grid renewables during an extended grid outage rather than repurposing their car’s power that they might also need to use to get around.

As for why it’s not “commonplace”, I think that’s pretty obvious. Teslas aren’t commonplace and those who can afford them can probably also afford a backup generator or battery system. Keeping the systems separate avoids unnecessary coupling of two systems that folks expect to be highly reliable; transportation and household power.


  • I’m assuming that the vehicle has zero value once the batteries start wearing out - obviously good maintenance and lighter use will extend lifecycle plus there could be scrap/resale value even at the EOL
  • I’m assuming that the taxes/incentives/maintenance/etc all roughly cancel each other out - this will vary a lot depending on location/circumstances/usage/etc

I’m honestly not sure what math you’re doing here, and I’m not sure it matters if there’s no feasible way to use the EV battery for standalone backup beyond tiny generator size outputs from the 12V system (1000-1500W sustained).

If you want to play with numbers, a small backup generator works out to around $0.75-$1.00/kWh delivered in gasoline costs - they’re not very efficient, even the little inverter ones. But you can get a used one for $300, and they work fine. I’m on my 5th year with a $300 used generator for my office.

Or go PHEV. :wink: I’ve got the best of both worlds, and if I hang a small inverter off the Volt’s 12V system, it’s a surprisingly good backup generator. 10kWh on battery, then a fairly efficient little gasoline generator with a 9 gallon tank plus whatever I have laying around the property, usually a minimum of 15 gallons, often closer to 30. I just don’t quite have that system wired up yet.

I honestly think there’s a weird social credit one gets for spending far more than what our house cost on a car, roof, and backup batteries. It ranges deeply into income circles I don’t play in and am not comfortable in.

1 Like

I wrote the following a week ago and didn’t post. Oops. But I second the idea of eeking by using a vehicle.

If you’re comfortable plugging in whatever you need into the SPS, then just do that for the sunny hours when the powers out. However, I don’t think there’s a way to use the SPS in any hardwired, subpanel configuration. It’s just a single outlet, so you can run an extension cord to your freezer or fridge. I think if you want a whole subpanel backed up, you’re gonna need batteries. Or fuggedaboutit vaporware iq8s from enphase?

I’m partial to the $150 inverter hooked up to an idling automobile to keep stuff at the barely adequate stage, but I haven’t endured what y’all in Texas did. Maintaining a separate generator is effort I don’t need. I’ve been meaning to test out whether the blower on my natgas furnace will start off an inverter, or whether I need to build a diy soft starter like this (sorry syonyk) video:

So my plan is to alternate between running the fridge, freezer, and furnace until power comes back. If it’s longer than a few days this is going to start sucking (and the gas tank will run out), but for a day or two, that’s manageable. If/when I get a SunnyBoy with an SPS, I’d do the same with that.

For you, I might suggest a power wall. They cost (too much) money, but throwing money at problems is sometimes the best option. My wannabe prepper MIL just spent something like $40k on a two PW system in northern Idaho. But turnkey is probable the right way to go for her (retiring soon physician).

First thing to do is work out what problem you want to solve. My requirement is being able to run fridge + Instant Pot to survive in the event of an extended outage (3-7 days). Riding out the occasional outage that is measured in hours, not days, is nice but not a strict requirement. A Powerwall works but is vastly more expensive than alternatives like SPS, generator, hybrid inverter + minimal battery. Those 2 PWs are good for maybe 10 years or so, and then she’ll need to replace them. I’m in my late 30s, and I can buy a generator and propane for life for less than just the initial $40K. You can even limp along indefinitely on SPS or hybrid inverter, which pay for themselves.

There is also this: This-Solar-Battery-Technology-Beats-Lithium-1.pdf ( The claim is that 10kWh of lead-carbon costs $2,500 and has similar performance to LiFePO4 in the real world.

Hard to tell what they’re talking about specifically, but the general concept of “Hey, don’t count lead acid out yet…” is quite correct.

Firefly Energy ( has a nice lead carbon foam AGM battery technology that seems to work about like LiFePO4, but it’s not cheap.

Any of the “Smart Carbon” or similar solar AGMs, treated properly, will work fine too, if you are dead set on not watering batteries.

The problem is that for the added “few times a year” annoyance of watering batteries, you can get amazing performance in flooded lead acid, plus the ability to ground-truth your state of charge and charging parameters from the specific gravity. The only downside is that they don’t like indefinite float service - you really want to be cycling them, or they are a bit prone to acid stratification (heavy stronger acid at the bottom, light weak acid at the top, and all sorts of bad things as a result of that).

Off Grid/RV Lead Acid Maintenance, Charging, and Failure Modes has probably more than you ever wanted to know about lead acid…

1 Like

My understanding is that you have an existing (~6kW) system but the inverter is NOT a Sunnyboy with SPS. And you’re considering replacing the inverter to get some sort of backup solution, be it offgrid (Outback) or grid-tied with SPS. And you maybe want to do it by powering a critical load subpanel. AirCon is a nice-to-have that requires 240V, but everything else (fridge, instant pot, natgas furnace?) is 120V.

Trying to get a week of service out of anything meant only for emergency backup is tough. What do you estimate your panels would have produced during the cold snap? 15kWh/day? Would having power only 4 hours/day be acceptable to you when it’s 4F outside? If the answer to that is no, you either need batteries (which don’t pay for themselves) or a $100-300 gas generator from Harbor Freight.

Or a Prius idling in the drive way with a suitably sized 12V inverter. I did that for ~20 hours, and burned 1.5 gallons of gas (it was running a very small load: sound system and oxygen concentrator). Idling burns about a quart an hour, but the prius shut the engine off for about a third of the time (after the battery is at an acceptable level again). 6 hours of extra idling wouldn’t make a difference to oil change intervals.

Using my van for that purpose would be 5x as wasteful, but I’d probably only need to run it a few hours per day. It’d be best if the grid came back within a day or two, but I’m sure you were expecting it back any hour now for 4-5 days.

Bit of a tangent, but if it’s 4 degrees fahrenheit outside a bit of manual labor can probably reduce the duty cycle of a refrigerator (and even possible a freezer) to near zero by allowing containers of water to freeze and putting them in the fridge periodically. Obviously, the effort cost of this would need to be weighed against the value of the food that might spoil and it would only make sense in certain circumstances. :stuck_out_tongue_closed_eyes:

Yes, you summarized things nicely. And our furnace is electric, unfortunately.

For reasons, we didn’t use the AC from Sunday morning on 2/14 until almost a month later. This is from before the outage, which hit us at 1:28 AM CT on 2/15. Since we survived the coldest storm on record with nothing more than blankets, I am not particularly concerned with keeping the house warm. We still had other options such as warming up in the car or huddling together in a sleeping bag at night.

We were lucky on food. Since they were “rotating” the blackouts, i.e. some people never lost power and others almost never had it, we were able to pull enough to reheat leftovers in the microwave and for my wife to cook one meal in ~15 mins. This is why my requirement is >1kW, which is roughly what both the microwave and Instant Pot pull. My C-max has a 150W inverter built in, but to cook I’d need to buy a larger inverter. This is still a good idea, and I might consider this as a backup for my backup.

I am more concerned about the possibility of losing power in the summer when the overnight low is in the 90s and daytime high >105F. SPS isn’t enough to run the AC, but the hybrid inverter should be able to run it for a good while. My current system hit ~2kW today around 10:30 AM. It seems reasonable to guess that a system twice as large would be able to run the AC continuously off of panels for 6~8 hours during the months when I would actually need it. We do occasionally have a cloudy day, but one day without AC is survivable.

Same idea, but my neighbor moved all of his food from the fridge/freezer into the garage.

When the ambient temperature is ~40F the fridge/freezer both work pretty well even without power. I did absolutely nothing, and my units didn’t even defrost.

If you’re concerned about summer, what’s the usual humidity in your area? 30-40 or less? A 50-60Ah SLA AGM with a smaller size 12v evaporative cooling setup blowing on you to manage to sleep might work. Definitely depends on your average humidity, if it’s 70+% there’s a serious efficiency loss.

For a larger, higher airflow evaporative cooler you might need to step up to a 80-100Ah battery. Coupled with rolling blackouts or a fair sized 12v backup/emergency solar panel, or 120vAC SPS → 12v charger and you can probably survive summer blackouts, even if it’s not as comfortable with full A/C system.

Doesn’t help with the fridge, but with a SB and SPS you probably can power it during the day, and if you have containers of water that you stick in the freezer side to get frozen during the day, put on the top shelf at night and just don’t open the door you might be able to keep your food well enough.