Battery storage paired with rooftop solar — a Tesla Powerwall 3, Enphase IQ Battery 5P, FranklinWH aPower, SolarEdge Home Battery, or similar — is the question every solar buyer in 2026 has to answer. Storage adds $10,000-18,000 to the install cost. The payback varies enormously by state, by utility tariff, and by whether the household values backup power independently of the financial return.
Three ways a battery earns money
- Time-of-use arbitrage. Charge the battery from solar at noon when the export credit is low, discharge in the evening when the import rate is high. Profit equals the spread times throughput.
- Self-consumption. Under net billing regimes where export credit is lower than import rate, every kWh you store and use yourself avoids the import rate. The avoided-cost savings is the spread.
- Demand-charge avoidance. Less common on residential tariffs, but some utilities are introducing residential demand charges. A battery that shaves the daily peak avoids the demand fee.
Backup power during outages is also a benefit, but it is not money. Households that experience frequent multi-hour outages may value backup independently of payback — this guide is about the financial return, not the resilience case.
The NEM 3.0 case study
California's 2023 transition to NEM 3.0 created the cleanest economic case for residential storage in the country. Under NEM 3.0, midday solar export earns 5-8¢/kWh, while evening import costs 35-55¢/kWh on time-of-use tariffs. The spread is 30-45¢/kWh, and a battery that cycles once a day pulling 10 kWh through it captures roughly $1,100-1,600/year in arbitrage value.
On a $12,000 battery, that is an 8-11 year simple payback before degradation. Within California IOU territory, storage now generally pencils out for households that consume meaningful evening electricity (heat pumps, EVs, late-night cooling).
The rest of the country
Outside NEM 3.0, the economic case for storage is weaker. In a state with full retail net metering, the export rate already equals the import rate, so the time-of-use spread the battery captures is zero. The only remaining value is demand-charge avoidance (rare) and backup (not money).
In net-billing states with smaller spreads (Massachusetts SMART, North Carolina Duke, Hawaii Customer Self-Supply), storage earns somewhere between the California case and the full-retail case. A 5-10¢/kWh spread on 3,500 kWh/year of throughput yields $175-350/year in savings — a 30-50 year payback on a $12,000 battery, which is not a real payback.
The IRA-era storage tax credit
Section 25D was extended to standalone battery storage (3 kWh or larger) in 2022, which made battery-only retrofits credit-eligible. That credit expired alongside the rest of 25D on December 31, 2025. Battery storage installed in 2026 does not receive a federal credit unless it is part of a commercial-owned system claiming Section 48E.
A handful of states still subsidize storage directly. California's Self-Generation Incentive Program (SGIP) pays per kWh of installed battery capacity, with higher rates for households in high-fire-risk zones. New York's NY-Sun program includes a storage adder. Maryland and Massachusetts have offered storage-specific incentives in past program years.
How to size for the right reason
The biggest sizing mistake homeowners make is buying a battery large enough to run the whole house for days. The economic-optimal size is small enough to cycle fully once per day on the highest-spread tariff hours. A 13-14 kWh battery is enough for most evening peaks; a second battery adds redundancy and longer outage coverage but adds little to the financial return because it does not increase daily cycles.
For backup-driven sizing, the math is different — you size to the loads you want to keep running, the duration you want, and your willingness to pay for that resilience independently of payback.
Bottom line
Buy a battery in California (NEM 3.0 makes the case), in Hawaii (high rates and self-supply), and in any household where evening electricity consumption is high and the local tariff has a real time-of-use spread. Outside those cases, the financial payback for storage in 2026 generally does not justify the cost — and the decision becomes about resilience, not return.