NEM 3.0 Battery Math: What an 8 kW System Actually Earns, With and Without a Battery

Why this article exists

Our NEM 3.0 explainer covers what California's Net Billing Tariff is, who it applies to, and how NEM 2.0 grandfathering works. This article does the part most guides skip: the actual math. We take a typical Southern California system — sized from our own 2025 install data — and run the numbers under NEM 2.0, under NEM 3.0 without a battery, and under NEM 3.0 with a battery, using verified 2026 SCE rates.

The system we model is not hypothetical. Across 1,299 completed Southern California projects in our published 2025 install data study, the median system size was 8.0 kW, 70% of projects included battery storage, and 31% of battery projects installed two or more units. That 70% battery attach rate is the market's answer to NEM 3.0 — this article shows the arithmetic behind it.

What actually changed: export value collapsed, import cost did not

Under NEM 2.0, every kilowatt-hour your panels exported earned a credit at close to the retail rate for that hour, minus small non-bypassable charges. Export a kilowatt-hour at noon, buy one back at night, and the trade was roughly even. The grid functioned like a free battery.

The CPUC's Decision D.22-12-056 ended that for anyone submitting an interconnection application on or after April 15, 2023. Under the Net Billing Tariff, exports are no longer credited at retail. SCE's Solar Billing Plan pays Energy Export Credits based on the CPUC's Avoided Cost Calculator — a value set hour by hour, differentiated across all 24 hours and by weekday versus weekend. For most daylight hours, when residential solar actually exports, that avoided-cost value works out to single-digit cents per kilowatt-hour. Customers who enroll before 2028 get their export credit schedule locked for the first nine years of operation, which at least makes the low value predictable.

Meanwhile, the price of the power you import kept climbing. Per the U.S. Energy Information Administration's Electric Power Monthly, California's average residential electricity price hit 35.25 cents per kilowatt-hour as of April 2026 — and SCE's time-of-use plans charge far more than that average in the evening. On SCE's TOU-D-4-9PM plan, summer on-peak power costs 58 cents per kilowatt-hour (about 48 cents after the baseline credit), while off-peak runs about 34 cents. That asymmetry — sell at roughly a nickel, buy at 34 to 58 cents — is the entire economic story of NEM 3.0.

The worked example: a median 8.0 kW SoCal system

Here is our model household, built from the median of our 2025 install data and typical SoCal conditions. All modeling assumptions are stated so you can check the arithmetic; your proposal from any installer should do the same.

Assumptions: an 8.0 kW system producing about 12,800 kWh per year (1,600 kWh per kW, a normal SoCal yield); household consumption of 9,600 kWh per year; SCE territory on TOU-D-4-9PM; without a battery, about 35% of solar production is consumed on-site as it is generated, and 65% is exported midday. For dollar rates we use 34 cents per kWh for off-peak retail, 48 cents for summer on-peak after the baseline credit (both from SCE's published TOU-D rates), a 40-cent blended year-round peak offset to account for lower winter rates, and a 5-cent average export credit consistent with avoided-cost-based Energy Export Credits for daylight hours. Real export schedules vary by hour and month; 5 cents is a reasonable planning average, not a tariff quote.

Reading the table: NEM 3.0 cut solar-only savings roughly in half

The same panels on the same roof earn about $4,185 a year under NEM 2.0 and about $1,940 under NEM 3.0 without storage — a drop of more than 50%. Nothing about the hardware changed. The only thing that changed is what the grid pays for the 65% of production the household could not use in the moment.

Add one battery and most of the loss comes back. A single Tesla Powerwall 3 stores 13.5 kWh; in our model it soaks up about 11 kWh of midday surplus and discharges it into the 4–9 p.m. window, when SCE power costs 48 to 58 cents in summer. Each shifted kilowatt-hour stops earning a nickel and starts avoiding roughly 40 cents blended across the year — a 35-cent-per-kWh swing, roughly $1,300 per year on about 3,740 shifted kilowatt-hours. Total savings land near $3,250, recovering about 78% of the NEM 2.0 economics while adding blackout backup that NEM 2.0 never provided.

This is also why 31% of our 2025 battery customers installed two or more units. A second battery extends evening coverage and whole-home backup, and for households with bigger evening loads — EV charging, air conditioning, pool equipment — there is more midday surplus to shift and more expensive evening consumption to offset. Battery-owning households can also switch to SCE's TOU-D-PRIME plan, where off-peak drops to about 26 cents while on-peak stays near 59 cents, widening the very spread the battery exploits.

Payback deltas: what the battery does to the timeline

Payback depends on price, and battery pricing is genuinely home-specific — panel condition, backup scope, and unit count move it, which is why we itemize Powerwall pricing in proposals rather than advertising a flat number. For a like-for-like comparison, assume a cash purchase of the 8 kW system at $24,000 and an incremental $14,000 for one installed Powerwall 3 — round modeling figures, not quotes.

Under NEM 2.0, the solar-only system paid back in roughly 5.7 years ($24,000 ÷ $4,185). Under NEM 3.0, solar alone stretches to about 12.4 years ($24,000 ÷ $1,940). Solar plus battery at $38,000 against $3,250 in annual savings comes to about 11.7 years — the battery pays for its own incremental cost and slightly shortens the payback, while adding backup power the solar-only system cannot provide. Rising rates compress all of these timelines: every SCE rate increase raises the value of self-consumed and shifted kilowatt-hours without changing what you paid.

One tax note that changes the 2026 math versus older articles: the federal Residential Clean Energy Credit (Section 25D) is not available for property placed in service after December 31, 2025, per the IRS. A purchased residential system installed in 2026 does not qualify for that 30% credit, and our modeled paybacks above assume no federal credit. Third-party-owned systems (lease or PPA) and commercial projects fall under separate business credit rules (Section 48E) with their own eligibility and deadlines — a reason some 2026 shoppers compare ownership structures more carefully than they would have in 2025. Income-qualified households may also access California SGIP equity incentives of up to $1,100 per kWh of battery storage, which can dramatically shorten battery payback where they apply.

Why battery attach is now the default, not the upsell

Put the pieces together and the 70% attach rate stops being surprising. NEM 3.0 pays single-digit cents for exports; SCE charges 34 to 58 cents for imports; a battery is the only device that converts the former into the latter. In our 2025 data the battery was not an accessory decision — it was the system design. Median size stayed a disciplined 8.0 kW rather than ballooning, because under net billing, oversizing an array to farm nickel-a-kilowatt-hour exports is bad math. The winning design pairs a right-sized array with enough storage to cover the evening peak.

OC Solar is a Tesla Powerwall Premier Certified solar and battery installer headquartered at 240 Progress, Suite 100, Irvine, California, serving Orange County, Los Angeles, San Diego, Riverside, San Bernardino, and Ventura counties with 30+ megawatts installed.

If you want the version of this math with your own roof, usage history, and rate plan instead of the median household's, request an estimate and we will run it line by line — the same way this article does. For the broader policy background, start with the NEM 3.0 explainer; for battery pricing structure, see the California battery cost guide.