Your utility bill arrives on a Tuesday. Last month it was $212. This month it's $238, and the only thing that changed is that you ran the dryer a bit more. That's the moment most people start googling home batteries.
And then the wall hits. Powerwall 3, Enphase IQ 5P, Franklin WH, Sonnen, Generac PWRcell, plus a dozen brands you've never heard of. Every installer swears theirs is the best. Every spec sheet uses different units. Nobody tells you what a residential solar battery storage system actually costs you per usable kilowatt-hour, which is the only number that matters.
I've had a Powerwall 3 on my own house since early 2024, and I've helped four neighbours and two family members spec their own systems since. I've also watched one of them get talked into a battery that was 40% oversized for their needs. So this isn't a spec-sheet roundup. It's what I've learned about comparing these things without getting taken for a ride.
Key takeaways
- Usable capacity, not nameplate capacity, is the number that determines what your battery can actually do. Most brands quote nameplate.
- Round-trip efficiency ranges from roughly 85% on older AC-coupled systems to about 97% on the newest DC-coupled units.
- Installed cost per usable kWh runs from around $750 on the low end to over $1,400 on premium high-cycle-life units, before any incentives.
- If blackout protection is your main motivation, you don't need solar at all. A standalone backup battery can do the job for less.
- In California, the economics changed dramatically after the net metering overhaul. A battery went from "nice to have" to "borderline mandatory" for new solar customers.
- Warranty throughput (MWh guaranteed over the warranty period) tells you more about battery quality than the year count.
How to compare residential solar battery storage options without drowning
Here's the thing nobody says out loud: most homeowners comparison-shop on three numbers—price, capacity in kWh, and warranty years. All three are the wrong starting point.
The right starting point is your evening and overnight load. Pull up your last 12 utility bills and find your average daily consumption. Then find what percentage happens after 4pm. For a typical family home with an EV, that's often 55–65% of total daily use. For a retired couple with gas heating and no EV, it can drop to 30%.
That percentage, multiplied by your daily kWh, tells you the minimum battery size that isn't a waste of money.
Usable vs nameplate capacity: the gap that ruins budgets
A battery rated at 13.5 kWh does not give you 13.5 kWh. Every lithium iron phosphate (LFP) unit limits depth of discharge (DoD) to protect cell life. Common DoD floors land between 90% and 100% for LFP, but older nickel manganese cobalt (NMC) cells often cap at 80%.
Then efficiency eats more. If you're charging from solar through a DC-coupled hybrid inverter, you'll lose around 3–5% on the round trip. On an AC-coupled system where the battery has its own inverter, that loss can reach 10–13%.
So a "13.5 kWh" battery on an AC-coupled setup might deliver 11.5 kWh usable on a good day. If your installer quotes you capacity without mentioning this, ask directly. The answer tells you whether they know their product.
AC-coupled vs DC-coupled: which one fits your situation
If you already have solar and it's staying on the roof, AC-coupling is usually the path of least resistance. The battery sits on the house side of your existing inverter, and retrofitting is straightforward.
If you're installing solar and battery together, DC-coupling through a hybrid inverter is generally more efficient. It shaves a few percentage points of loss off every cycle—over 10 years and 3,500 cycles, that adds up to real money.
What surprised me: DC-coupled systems used to be a premium niche. In 2024 and 2025, hybrid inverters like the Enphase IQ8 and SolarEdge Home Hub closed most of the price gap. Now the difference is often $400–$700 on a full install, which the efficiency gain usually pays back in four to six years.
What does a home battery actually cost in 2026?
Cost is where comparison articles get vague, so let's be specific. These are numbers from quotes I've seen or gathered from installers in the last 18 months, all before any federal or state incentive.
| System | Usable capacity | Continuous power | Round-trip efficiency | Warranty throughput | Installed cost (approx.) |
|---|---|---|---|---|---|
| Tesla Powerwall 3 | 13.5 kWh | 11.5 kW | ~97% (DC-coupled) | ~37 MWh | $11,500–$14,000 |
| Enphase IQ Battery 5P | 5.0 kWh per unit | 3.84 kW per unit | ~90% (AC-coupled) | ~15 years / 6 MWh per unit | $6,500–$8,000 per unit |
| Franklin WH aPower | 13.6 kWh | 10 kW continuous, 15 kW peak | ~92% | ~30 MWh | $12,000–$15,000 |
| Sonnen ecoLinx | 10–20 kWh (configurable) | 8 kW | ~94% | 10,000 cycles | $16,000–$24,000 |
| Generac PWRcell 2 | 9–18 kWh (modular) | 4.5–9 kW | ~96% | ~10 years / varies by module | $10,000–$17,000 |
Notice something? Cost per usable kWh swings from around $850 on a Powerwall to over $1,600 on a Sonnen. Both are good batteries. The Sonnen is built for higher cycle life and has a longer designed service window, which matters if you plan to stay in the house for 15+ years.
If you're likely to move in 7 years, the Sonnen premium is hard to justify. That's a personal call, and I'd argue it's more important than any spec.
Is a Tesla Powerwall 3 still the right choice?
For most people in the US, yes—with caveats.
The Powerwall 3 is the highest-volume residential battery on the market, which means installers know it inside out, replacement parts are widely stocked, and pricing is competitive because Tesla's own installer network keeps pressure on third parties. Its integrated inverter also means you don't need a separate solar inverter, which simplifies the whole system.
The caveats matter, though.
- You're buying into Tesla's ecosystem. If service support in your region is thin, that's a real problem during a warranty claim.
- The integrated inverter limits you to specific panel configurations. If your roof has complex shading or multiple orientations, some installers prefer a separate inverter and AC-coupled battery.
- Software updates come from Tesla on Tesla's schedule. You can't tune the dispatch algorithm yourself.
Franklin WH wins on one specific thing: generator integration. If you already have a whole-home generator and want the battery to work alongside it, Franklin's aPower handles that handoff more gracefully than Powerwall. I've seen two installs where this mattered.
Best home battery backup without solar
You can install a battery with no solar at all. It charges from the grid during off-peak hours and discharges during peak hours or outages. This is called grid-charged backup, and it's often overlooked.
Two systems dominate this category. The Tesla Powerwall 3 works fine without solar—it just charges from the grid. And the Generac PWRcell 2 was designed with this use case in mind, especially for homes that already have a Generac generator or want one for extended outages.
Cost is roughly the same as a solar-paired install, minus the panels. For a homeowner whose utility offers generous time-of-use arbitrage (peak rates at 4–9pm at 2–3x the off-peak rate), a grid-charged battery can pay back in eight to eleven years without any solar at all. That's a longer payback than a solar+battery combo, but it's a real option if your roof is shaded or leased.
How many batteries do I need?
Work backwards from your critical loads, not your total consumption.
Make a list of what you'd genuinely want running during a multi-day outage: fridge, freezer, a few lights, internet router, phone chargers, sump pump or well pump if you have one, and one room's worth of heating or cooling. For most homes, that's 3–6 kWh per day.
A single 13.5 kWh battery covers that for two full days. If you want to run the whole house as normal—including the dryer, oven, and A/C—you're looking at three to five batteries, and the economics stop making sense for most people.
My honest recommendation: size for critical loads plus one luxury item (electric kettle, coffee maker, or a space heater in winter). Everything beyond that is emotional, not practical.
California and the net metering shift
If you're in California, the rules are different and the stakes are higher.
Since the switch to net billing (NBT) for new solar customers, exported power earns far less than it did under the old net energy metering regime. What used to be a 1:1 credit for every kWh you sent to the grid now lands somewhere in the range of $0.05–$0.08 per kWh during peak solar hours, against retail rates that run $0.30–$0.55 per kWh depending on your utility and tier.
The math is blunt: without a battery, you're selling cheap and buying expensive. With a battery, you store the cheap solar and use it during expensive evening hours, which is the entire point.
Three other states now have similar structures worth checking before you buy: Hawaii (which has had battery-first rules for years), Illinois, and Maryland. If your utility offers a time-of-use rate with a peak window, a battery almost always makes sense on paper. If it doesn't, run the numbers carefully.
The one number nobody quotes you
Ask your installer this question: "What's the round-trip efficiency measured at the meter, not at the battery?"
Most won't know. The ones who do are usually the ones worth hiring.
Because that single figure—measured at your utility meter, through the whole system, on a typical day—determines what percentage of your solar production survives the round trip through storage. A 90% efficient system loses one kWh out of every ten you put in. A 97% system loses three out of a hundred. Over 15 years and 4,000 cycles, on a 13.5 kWh battery, that's a difference of roughly 1,500 kWh of usable energy—about a full year of partial self-consumption for a small home.
Which brings up the second question nobody asks: throughput warranty. A battery guaranteed for 10 years but capped at 20 MWh of throughput might hit that ceiling in year seven if you cycle it hard. A battery guaranteed for 10 years at 37 MWh gives you real headroom. Read the fine print. The year count is marketing; the MWh number is engineering.
One more thing worth checking: does your utility have a demand response program that pays you for letting them dispatch your battery during grid stress events? In California, these programs exist and pay real money—a few hundred dollars a year in a good season. Your installer probably won't bring it up unless you ask.
What I'd do differently
If I were speccing my system again today, I'd spend more time on three things and less on brand comparisons.
First, I'd model my actual hourly usage using the data from my smart meter, not a daily total. The daily total hides the shape of the day, and the shape is what determines whether a battery helps you.
Second, I'd ask every installer for a reference customer within 30 miles who's had the same system for at least two years. Then I'd call them. Two of the four neighbours I helped got very different answers from that phone call than from the sales pitch.
Third, I'd ignore any comparison that lists specs without a cost-per-usable-kWh column. Specs are easy to manipulate. The cost per usable kWh, measured over the full warranty window, is the only number that compares fairly across brands.
Which battery you choose matters less than whether the system is sized to your actual evening load, installed by someone who'll still answer the phone in year four, and matched to whatever your utility is actually paying for right now. Get those three right and the hardware almost sorts itself out.
