
Home Battery Guide
Does a Home Battery Work Without Solar? Grid-Charged Backup
Grid charge rates, refill times, runtime with no recharge, and the state programs that pay for a standalone battery. Checked August 16, 2026.
Quick answer: Yes. A grid-charged battery with no solar attached is a working backup system, and for the outage pattern most US homes actually see it is enough: EIA reports the average US customer lost 11 hours of power across all of 2024, and about 2 of those hours came from something other than a major storm. One 13.5 to 15 kWh battery covers that outright. Solar earns its place when your risk is the multi-day tail, because a battery with no generation on site gets exactly one tank and then waits for the grid. The number to ask for is the charge rate: Powerwall 3 takes 5 kW AC on a single unit, FranklinWH aPower 2 takes 8 kW, and that decides how fast you refill when the power flickers back.
Best for
US homeowners deciding whether to buy a battery alone or wait and buy solar with it.
Wrong fit
Off-grid homes, RV and marine systems, or homes with no grid service at all.
Tradeoff
A grid-charged battery is cheaper and covers a normal outage year, but it cannot make energy, so every hour past its one tank depends on the grid coming back.
Yes. A grid-charged battery with no solar attached is a working backup system, and for the outage pattern most US homes actually see it is enough: EIA reports the average US customer lost 11 hours of power across all of 2024, and about 2 of those hours came from something other than a major storm. One 13.5 to 15 kWh battery covers that outright. Solar earns its place when your risk is the multi-day tail, because a battery with no generation on site gets exactly one tank and then waits for the grid. The number to ask for is the charge rate: Powerwall 3 takes 5 kW AC on a single unit, FranklinWH aPower 2 takes 8 kW, and that decides how fast you refill when the power flickers back.
Almost every page that answers this question is published by someone who sells solar. That is not a conspiracy, it is just who writes about batteries. It does mean the honest version of the answer, the one with recharge rates and outage-hour data in it, is hard to find. So here it is with the numbers attached.
How we checked this
Last updated: August 16, 2026. Every figure below was pulled or re-pulled on that date.
- Capacity, power and charge rates come from the manufacturers' own current datasheets, not product pages or installer blogs. Charge power is the number that decides the no-solar case, so we quote it per model and name the document it came from.
- Outage data comes from the EIA's Electric Power Annual 2024, published December 1, 2025. It is the only national reliability dataset with a per-customer hour count.
- Incentives come from the program administrator's own page, with the amount, the class it applies to, and the date we checked it. Where a program page could not be reached, we say so rather than repeating a number we did not re-verify.
- Installed costs are our own mid-2026 planning benchmarks, cross-checked against EnergySage's published averages with their update dates.
- What we did not do: we did not bench-test any of these systems, and we do not publish our own measured runtime or pricing dataset. Every number here is either quoted from a primary document or arithmetic on those quoted numbers, and the arithmetic is shown.
Battery only vs battery plus solar
| Decision input | Battery only, grid-charged | Battery plus new solar |
|---|---|---|
| Gross installed cost, our mid-2026 benchmark | $11,500-$18,000 for one whole-home battery | $11,500-$18,000 plus the solar project |
| National average, EnergySage | About $15,650 for a typical 13.5 kWh battery (first-half-2026 data, page updated July 13, 2026) | About $31,135 for a 12 kW system at $2.60 per watt (updated June 30, 2026), so about $46,785 combined |
| Benchmark to sanity-check the quote | About $1,040 per usable kWh is normal; near $1,500 per usable kWh needs an explanation | Same battery benchmark, plus dollars per watt on the solar half, priced separately |
| Recharge while the grid is down | 0 kWh. The battery holds one tank | Whatever the array makes that day, capped by the battery's charge rating |
| Refill once the grid returns | 5 kW AC on one Powerwall 3, 8 kW on a FranklinWH aPower 2 | Same, plus daytime solar |
| Federal tax credit for a 2026 install | $0. Section 25D is not available for property placed in service after December 31, 2025 | $0, same rule, same date |
| State and utility programs | CT, NH, MA and RI all accept a standalone battery. See the table below | Same programs, same amounts |
| Main bill-savings lever | The time-of-use spread between charging and discharging, plus program payments | Offsetting retail imports with self-generated power, plus export credit |
| Where it fails | A multi-day outage past the first tank | Nothing about backup. It fails on budget and on roof suitability |
The battery half of the cost is the same in both columns. That is the point people miss when they ask whether they need solar: solar is not a discount on the battery, it is a second project with its own price tag.
Grid charging, model by model
Discharge power is on every spec sheet and in every sales deck. Charge power is buried, and when there is no solar it is the only way energy gets into the box. Here is what the three most-quoted systems actually publish.
| Specification | Tesla Powerwall 3 | Enphase IQ Battery 5P | FranklinWH aPower 2 |
|---|---|---|---|
| Model number on the datasheet | 1707000-xx-y | IQBATTERY-5P-1P-NA | APR-10K15V2-US |
| Usable energy per unit | 13.5 kWh AC nominal | 5.0 kWh | 15 kWh AC |
| Continuous discharge | 5.8, 7.6, 10 or 11.5 kW AC depending on the configured rating | 3.84 kVA at 240 VAC, 3.33 kVA at 208 VAC | 5, 7.6, 9.6 or 10 kW, up to 11.5 kVA |
| Maximum continuous charge power | 20.8 A AC / 5 kW for a single unit; 33.3 A AC / 8 kW with up to three Expansion units | Not published as a separate rating | 8 kW continuous |
| Motor start capability | 185 LRA | Up to 48 A LRA per unit | 185 A LRA |
| Peak output | Not published as a discrete peak figure | 7.68 kVA for 3 seconds, 6.14 kW for 10 seconds | 15 kW for 10 seconds, 25 kW transient for 1 second |
| Round-trip efficiency | 89% solar to battery to home or grid | 90% AC round trip, 96% DC | 90% grid to battery to load |
| Backup controller required | Gateway 3, Backup Switch or Backup Gateway 2 | IQ System Controller 3/3G | aGate |
| Published warranty | 10 years | Up to 15 years or 6,000 cycles, above 60% capacity | 15 years or 60 MWh throughput |
| Hours to refill from empty at the published charge rate | 2.7 h at 5 kW | 1.3 h per unit at 3.84 kW | 1.9 h at 8 kW |
| The catch when there is no solar on site | The 15.4 kW off-grid discharge boost is listed PV Only, so a grid-only install stays at its on-grid rating | Backup is not supported on 208 V single-phase | Charge ceiling tracks the configured rating: 5 kW and 7.6 kW units cap input at 5 kW and 7.6 kW |
Notes on that table, because two cells need them. Enphase does not publish a separate maximum charge rating for the 5P in its North America datasheet, so the 1.3 hour refill is our arithmetic on the unit's single 3.84 kVA continuous rating, not a published figure. And every refill time above is capacity divided by charge power. Real charging tapers near full and loses roughly 10 percent to conversion, so treat these as floors, not promises.
Tesla Powerwall 3. It has an integrated solar inverter rated to 20 kW DC of PV input, which is why people assume it needs panels. The datasheet also publishes an AC charge rating, 20.8 A and 5 kW for a single cabinet, and that is the path grid charging uses. What the two numbers give you is a lopsided ratio: 11.5 kW out, 5 kW in. In a rolling outage that is the constraint that bites, because the grid may only be up for an hour at a time. Adding Expansion units lifts AC charging to 8 kW and adds energy, not inverter output.
Enphase IQ Battery 5P. The 5 kWh module is the smallest sensible unit here, and three of them get you to 15 kWh with 11.52 kVA of aggregate output, which also means 11.5 kW of aggregate charge capability. Its datasheet lists "NEM integrity" among the system's services, which is a quiet acknowledgment that grid charging and net metering rules interact and the equipment is expected to police it. Ask your installer what that setting will be on your tariff.
FranklinWH aPower 2. The best charge-to-capacity ratio of the three: 8 kW into 15 kWh. FranklinWH's own datasheet describes storing "grid energy during lower rate periods" as a design intent, not a workaround. The aGate is required and carries the control cost, so the first cabinet is the expensive one.
For the full three-way comparison on price, expansion and warranty, see Powerwall 3 vs Enphase vs FranklinWH.
What happens on day two and day three
With no solar, runtime is one division problem. Usable energy divided by the load you actually back up.
| Continuous backup load | 13.5 kWh, no recharge | 15 kWh, no recharge | Two 13.5 kWh units, no recharge |
|---|---|---|---|
| 0.5 kW | 27.0 h | 30.0 h | 54.0 h |
| 1 kW | 13.5 h | 15.0 h | 27.0 h |
| 2 kW | 6.8 h | 7.5 h | 13.5 h |
| 3 kW | 4.5 h | 5.0 h | 9.0 h |
| 5 kW | 2.7 h | 3.0 h | 5.4 h |
These are load levels, not appliance lists, and they are gross of conversion losses. Shave roughly 10 percent off each cell. Enphase publishes 90 percent AC round trip and FranklinWH publishes 90 percent grid to battery to load, while Tesla's 89 percent figure is measured on a solar-shifting path rather than a grid round trip, so treat it as the closest published proxy rather than an exact match. Which of your circuits lands at 0.5 kW and which at 5 kW is the whole sizing question, and it is worked through in the runtime guide and the kWh sizing guide.
Now the part that only matters when there is no solar. Most outages are not one clean 72-hour block. Power comes back, drops again, comes back. Every hour the grid is up is an hour you are refilling, and the charge rate decides how much you bank.
| One hour of restored grid buys you | At a 0.5 kW load | At a 2 kW load | At a 5 kW load |
|---|---|---|---|
| Powerwall 3, single unit, 5 kW charge | 10.0 h | 2.5 h | 1.0 h |
| Powerwall 3 with Expansion units, 8 kW charge | 16.0 h | 4.0 h | 1.6 h |
| FranklinWH aPower 2, 8 kW charge | 16.0 h | 4.0 h | 1.6 h |
| Three Enphase 5P units, 11.5 kW aggregate | 23.0 h | 5.8 h | 2.3 h |
That table is arithmetic on the published charge ratings, again gross of losses and taper. It is also the single strongest argument for a grid-charged battery, and no solar page will show it to you.
The outage numbers behind the decision
EIA's Electric Power Annual 2024, published December 1, 2025, is the reference point worth arguing from.
| What EIA measured, 2024 | Hours per average US customer |
|---|---|
| Total electricity interruption | 11 h |
| Attributed to major events | About 9 h |
| Everything else, the routine stuff | About 2 h |
| Major-event average, 2014 through 2023 | Nearly 4 h per year |
| Worst state in 2024, South Carolina | Nearly 53 h |
| Arizona, South Dakota, North Dakota, Massachusetts | Under 2 h each |
Read the first and last rows together. If your county looks like the national picture, one grid-charged battery is not a compromise, it is oversized for the job and you are buying it for the tail risk. If you live where the 2024 number was 53 hours, the tail risk is your normal, and a battery with no way to refill itself is a different proposition entirely.
Programs that pay for a battery with no solar
Standalone storage is not a loophole in these programs. Several of them were written for it.
| Program | Where | What it pays | Solar required | Checked |
|---|---|---|---|---|
| Energy Storage Solutions | CT, Eversource and United Illuminating | Enrollment $30/kWh standard, $130/kWh grid edge. Performance years 1-5: $300/kW-year standard, $450 underserved, $550 low income | No | Aug 16, 2026, rates effective April 1, 2026 |
| NH Clean Energy Fund battery storage | NH, Eversource | $230 per eligible kWh, capped $3,000 per account | No. Eversource states solar is not mandatory | Aug 16, 2026 |
| ConnectedSolutions battery | MA, National Grid | $275/kW of average summer performance, June 1 to September 30, up to 60 events, 3 h maximum each. National Grid cites about $1,200 average per year | No. "Customers who don't have a renewable energy system but do have an energy storage system" may participate | Aug 16, 2026 |
| Energy Storage Rebate | RI, Office of Energy Resources | $250/kWh residential capped $8,000; income-eligible $500/kWh capped $16,000 | No | Not re-verified Aug 16, 2026. See the note below |
| SGIP | CA, PG&E, SCE, SoCalGas, SDG&E | CPUC's participation page lists $850/kWh Equity and $1,000/kWh Equity Resilience | Not stated on the page we checked | Aug 16, 2026, with a caveat below |
| Section 25D federal credit | US | $0 | Not applicable | Aug 16, 2026 |
The federal row is not a typo. The IRS states plainly that the Residential Clean Energy Credit "is not available for any property placed in service after December 31, 2025," and its Public Law 119-21 fact sheet, FS-2025-05 issued August 21, 2025, adds that the expenditure counts when the original installation is completed, not when you paid. Any 2026 proposal, with solar or without, that still shows a 30 percent federal credit in the payback math is working from a dead number. That is covered in full in the 2026 battery tax credit guide.
Program detail lives on the state pages: Connecticut, New Hampshire, Massachusetts, Rhode Island and California SGIP.
When adding solar is worth it
Solar changes three things, and only three.
It refills the battery mid-outage. This is the real one. Past the first tank, a grid-charged battery is waiting and an array is working. If your outage risk is hurricanes, ice storms or wildfire de-energization, that difference is the decision. If your risk is a substation fault that clears in ninety minutes, it is not.
It changes what the battery can save you. Without generation on site there is nothing to export, so export credit is not part of your math at all. Your only savings lever is the time-of-use spread: charge when power is cheap, discharge when it is expensive, and keep the difference minus about 10 percent in round-trip losses. That is real money on a steep California or New England rate, and it is close to nothing on a flat rate. Run your own tariff before assuming either.
It adds a tariff question to answer in writing. If you already have solar under net metering or California's net billing tariff, adding a battery that charges from the grid can change how your export credits are calculated, and some programs limit grid charging outright. Enphase building "NEM integrity" into the 5P's feature list tells you this is a live constraint, not a theoretical one. Ask the installer to state, in writing, which tariff your account will be on after the battery is energized and whether grid charging is permitted on it. The NEM 3.0 guide covers the California version of this.
What solar does not do is make the battery cheaper, faster to charge from the grid, or better at starting a motor. Those are hardware numbers and they are the same in both columns.
What we could not verify
Publishing the gaps is part of the method.
- Tesla's own servers refused our fetch on August 16, 2026. Both energylibrary.tesla.com and tesla.com returned HTTP 403. The Powerwall 3 figures above come from the 2025 Powerwall 3 datasheet PDF as republished by a Tesla-certified installer, which matches Tesla's official document. Both URLs are in the sources. Confirm against the current datasheet at your quote date.
- Enphase does not publish a maximum charge power for the IQ Battery 5P. The datasheet gives one continuous rating, 3.84 kVA. Our refill figure is derived from it, and it is labeled derived in the table.
- Rhode Island's rebate figures were not re-verified today. energy.ri.gov returned HTTP 403 to our fetcher. The $250/kWh and $8,000 cap come from our own Rhode Island page, checked July 15, 2026. Treat them as one month stale, not confirmed today.
- SGIP's current general-market residential step is not published with a 2026 date on the CPUC participation page we read. The same page's budget-status text still refers to a December 2020 update. We are quoting what the page says, not asserting that those steps are open. Confirm with your program administrator before it goes in a quote.
- Daily solar production during an outage is not a number we will invent. It depends on array size, season, weather and shading, and it swings by a factor of three across those. Model your own address with NREL PVWatts before you let a proposal claim a specific recharge rate.
- Whether a specific utility tariff restricts grid charging on your account is not published in any single national source we could find. It is an installer question with a written answer.
Quote red flags
- The proposal quotes discharge power and never states the charge rate, so you cannot tell how fast the system recovers between outages.
- The installer says a battery "needs" solar to qualify for an incentive without naming the program and the eligibility line that says so.
- A 2026 quote showing a 30 percent federal credit for a cash or loan purchase.
- A payback figure built on export credit for a system with no generation on site.
- A recharge or resilience claim that assumes a specific number of solar kWh per day with no location, season or array size behind it.
- Solar and storage bundled into one total, so the battery cannot be benchmarked in dollars per usable kWh on its own.
What to ask before signing
- What is the maximum continuous charge power of this system in kW, and does it change with the configured output rating?
- If the grid is down for three days, what is the runtime at my actual backup load, and what recharges it?
- Which incentive are you claiming, at what amount, and what is the effective date of that number?
- Is the battery priced separately from any solar, with its own dollars per usable kWh?
- Which tariff will my account be on after this is energized, and is grid charging permitted on it?
Holding a proposal already? Run it through the quote calculator or send it in for a quote review. We do not sell batteries and we do not auction contact details, so the answer comes back straight.
Where this fits
If you have not set a budget yet, start with the real cost of a home battery. If backup scope is still open, decide whole-home or critical loads before anyone quotes you. If a generator is still on the table, the battery vs generator comparison is the other half of this decision. And if you are shortlisting hardware, the best home batteries roundup and the warranty guide are the next two stops.
Commercial note
Home Battery Guide may earn from affiliate links or flat-fee referrals to named vetted installers. Rankings do not move with compensation. We do not sell the same lead to multiple installers, and a referred installer quote still has to pass the same quote check.
Frequently Asked Questions
Can a home battery charge from the grid?
Yes, and the manufacturers publish the rate. A single Tesla Powerwall 3 accepts 20.8 A AC, which is 5 kW, rising to 33.3 A and 8 kW with Expansion units. A FranklinWH aPower 2 accepts 8 kW continuous. Enphase publishes one 3.84 kVA continuous rating per IQ Battery 5P unit rather than a separate charge figure. Whether your utility tariff permits grid charging is a separate question to answer in writing.
How long does a battery last in an outage with no solar?
Usable energy divided by your backup load, minus about 10 percent for conversion losses. A 13.5 kWh battery runs 27 hours at 0.5 kW, 6.8 hours at 2 kW and 2.7 hours at 5 kW before losses. With no generation on site there is no recharge until the grid returns.
Do I need solar to get a battery incentive?
Not in the New England programs. Connecticut's Energy Storage Solutions, New Hampshire's Clean Energy Fund battery program and Massachusetts ConnectedSolutions all accept a standalone battery, and National Grid states outright that customers without a renewable energy system may participate with storage alone. The federal Section 25D credit is a different matter: it is unavailable for property placed in service after December 31, 2025, with or without solar.
Is a battery without solar worth it?
It depends on which outage you are buying against. EIA measured 11 hours of interruption for the average US customer across all of 2024, and about 2 of those hours were routine rather than storm-driven. One battery covers that with room to spare. If your state looks more like South Carolina's nearly 53 hours in 2024, the multi-day tail is your real exposure and the absence of on-site generation is a genuine limitation.
Can I add solar to a grid-charged battery later?
Usually, and the architecture decides how cleanly. Powerwall 3 has an integrated solar inverter rated to 20 kW DC of PV input, so panels connect to the battery itself. Enphase and FranklinWH are AC-coupled, so solar arrives through its own inverter and the battery does not care. Ask now, in writing, what adding an array later would cost in equipment and labor, because that answer belongs in today's comparison.
Does grid charging hurt the warranty?
It is normal use rather than an exception, but every warranty is capped by cycles or throughput as well as years, and heavier daily cycling reaches those caps sooner. Enphase caps the 5P at 6,000 cycles inside its 15 years. FranklinWH caps the aPower 2 at 60 MWh of throughput. Tesla publishes 10 years. If you plan aggressive daily time-of-use cycling, ask how many cycles per year the installer is modelling. The warranty guide has the full cross-brand table.
Sources
- Tesla Powerwall 3 datasheet, official Energy Library URL
- Tesla Powerwall 3 datasheet, 2025 edition as republished by a certified installer
- Enphase IQ Battery 5P data sheet, DSH-00010-11.0-EN-2025-05-06
- Enphase IQ Battery 5P product page
- FranklinWH aPower 2 datasheet, SKU APR-10K15V2-US
- EIA: Hurricanes in 2024 led to the most hours without power in the United States in 10 years, December 1, 2025
- IRS: Residential Clean Energy Credit
- IRS FS-2025-05: FAQs on Public Law 119-21 modifications to sections 25C, 25D and others
- Connecticut Energy Storage Solutions: program changes effective April 1, 2026
- Eversource: New Hampshire Clean Energy Fund home battery incentive
- National Grid Massachusetts: ConnectedSolutions battery program
- CPUC: Participating in the Self-Generation Incentive Program
- EnergySage: how much do home batteries cost
- EnergySage: solar panel cost data
- NREL PVWatts Calculator
Methodology
These guides are built from public specifications, primary program pages, utility documentation, manufacturer materials, and repeated buyer questions that show up in quote and installation decisions.
Manufacturer and installer responses can clarify pricing bands, warranty terms, support footprint, and common mistakes. They do not move a page up the shortlist on their own.
Next Step
What to do next
Use one of these three paths. They are here to move the decision forward, not add more noise.
Want the full buyer path in your inbox? We send the short version.
Related Guides
home battery • us
Home Battery Warranty Guide: What Each Brand Actually Covers
Compare Tesla, Enphase, FranklinWH, Generac, SolarEdge, Panasonic, and sonnen warranty terms, coverage exclusions, and installer-failure risk.
home battery • us
Battery Tax Credit 2026: What Still Applies
The 2026 home battery incentive guide, with federal credit caveats, lease/PPA questions, and state program sources.
home battery • us
Best Home Batteries for Serious Buyers
A buyer-first shortlist of the home batteries worth comparing in 2026, with fit, tradeoffs, and quote questions.
home battery • texas
Battery vs Generator in Texas
A Texas buyer guide to choosing between a home battery, generator, or hybrid backup setup.
home battery • us
Best Portable Power Stations for Backup
When a portable power station is enough, when it is not, and how to compare EcoFlow, Anker, Bluetti, Jackery, and Goal Zero.
home battery • new england
ConnectedSolutions Battery Program
How the ConnectedSolutions battery program works, what it can pay, and what New England battery buyers should ask.