6.5¢ vs 33.6¢: California Net Metering Rates Under NEM 3.0

By SolarCostData Data Desk · Reviewed September 2026 · Corrections log

If you’re a California homeowner pricing solar in 2026, NEM 3.0 is the rule that decides whether it pays — more than the panel brand, more than the installer, arguably more than the (now $0) federal credit. It’s widely misunderstood, so here’s what actually changed and what to do about it.

NEM 2.0 vs. NEM 3.0: what actually changed

Under the old NEM 2.0, when your panels sent surplus power to the grid, the utility credited you at roughly the retail rate — close to 30¢/kWh, and in some peak periods higher. Export a kilowatt-hour at noon, pull one back at 8 p.m., and it was nearly a wash. That’s what made California solar so lucrative for a decade.

The Net Billing Tariff — NEM 3.0 — ended that. Since April 15, 2023, exported solar is credited at an avoided-cost rate instead: what it’s actually worth to the grid at that exact moment. On average that’s about 5–8¢/kWh, roughly a 75% cut from retail. And it’s not one number — the tariff has hundreds of hourly export values that swing by season and time of day, spiking on hot summer evenings when the grid is stressed.

The practical effect is a complete inversion of the old strategy. Overproducing at midday and leaning on the grid at night used to be fine. Now it’s the worst thing you can do, because you’re selling cheap and buying expensive.

Put the two numbers next to each other and the change stops being abstract. A California household still buys grid power at roughly 33.6¢/kWh. Because the ACC prices every hour separately, there is no single export rate to quote. Commonly reported annual averages fall in the 5–8¢/kWh band, and this site models California at 6.5¢, the midpoint — roughly a fifth of what it costs to pull that same kilowatt-hour back after dark. Your own average depends on when your array actually exports, since an evening kilowatt-hour is worth far more than a midday one.

NEM 3.0 export rates: the 576-number tariff

The single biggest misconception about NEM 3.0 is that there’s “an export rate.” There isn’t. Export credits come from California’s Avoided Cost Calculator (ACC), which publishes a different value for every hour of the day, every month of the year, split between weekdays and weekends — 576 distinct prices a year. Each one reflects what the grid would otherwise have paid to source that power at that instant.

That structure is the whole strategy in a nutshell:

• Midday (roughly 9 a.m.–3 p.m.): solar is everywhere, the grid needs it least, and export values sit at the bottom of the range — often that ~5–8¢.
• Summer evenings (the 4–9 p.m. window): demand surges as solar fades, and export values climb, sometimes into the low-to-mid teens of cents per kWh.

So a kilowatt-hour you can hold and export at 7 p.m. on a hot August day is worth multiples of the same kilowatt-hour dumped to the grid at noon. This is why NEM 3.0 rewards batteries that discharge into the evening far more than it rewards oversized arrays that just push midday surplus.

Why the answer is usually a battery

The whole game under NEM 3.0 is the spread between two numbers: what California utilities charge for power on a summer evening and the few cents you’d get for exporting at midday. On the big three utilities, peak time-of-use rates land in the 40–70¢/kWh range during that 4–9 p.m. window — the average all-in residential rate at PG&E was already around 39¢/kWh in early 2026, and SCE’s default time-of-use plan pushes summer peak pricing toward 50–55¢/kWh. Every kilowatt-hour you can store at midday and use yourself in the evening is worth that difference instead of the tiny export credit .

That’s why solar-plus-battery, not solar alone, is the 2026 California default. A well-designed pairing in a high-rate territory like SDG&E can offset roughly $2,500–$4,000 of electricity in the first year and land a cash payback around 7–11 years — while solar-only under net billing stretches longer because so much of the production gets exported for pennies. Run your own spread in the battery worth-it calculator ; it’s built around exactly this peak-rate-minus-export-credit math.

Sizing under net billing: aim for self-use, not surplus

NEM 3.0 quietly rewrites the sizing rulebook too. Under NEM 2.0, oversizing an array to bank a big export surplus made sense, because that surplus was worth almost retail. Under net billing it isn’t. Bill credits still offset your monthly charges, but if you end your 12-month cycle with a net surplus, the utility settles it at a low Net Surplus Compensation rate — a market-based figure that’s usually only a couple of cents per kWh. Overbuilding to export is close to donating power.

The winning design is different: enough panels to run the house and charge a battery for the evening, with the battery sized to carry you through the pricey 4–9 p.m. peak. If you have an EV or plan to electrify heating, factor that load in up front — it’s self-consumption you can shift to cheap hours, which is worth far more than any export credit. Some utilities now offer super-off-peak daytime windows (SDG&E expanded a weekday 10 a.m.–2 p.m. super-off-peak in 2026), which changes when it’s smart to charge cars and batteries from the grid versus from your own roof.

What the battery actually costs — and why the spread still wins

Recommending a battery is easy; paying for one is the real question, so be clear-eyed about the number. In 2026 a single home battery of around 10–13.5 kWh runs roughly $10,000–$16,000 fully installed before any incentive — call it $850–$950 per usable kWh across the leading brands . And here’s the accuracy point most 2026 sales decks blur: a cash or loan buyer gets no federal tax credit on that battery. The 30% residential credit ended December 31, 2025, so the old habit of quoting battery cost “after the 30% credit” no longer applies to a purchase.

So the case for storage under NEM 3.0 has to stand on the spread, not a tax rebate. It usually does: if a battery lets you avoid buying 5–8 kWh of 50¢+ evening grid power every day instead of exporting it for pennies, that’s a few dollars a day — often $700–$1,200 a year of avoided peak cost — chipping away at that install price. In 2026 that spread has to carry the whole case on its own, because the SGIP rebate is closed and the federal credit is $0. The battery earns its keep by self-consumption arbitrage, which is exactly why NEM 3.0 makes storage pencil out in California when weak export credits alone wouldn’t. Model your own version of that trade in the battery worth-it calculator linked above rather than trusting a flat “batteries pay for themselves” claim; the national battery guide covers what those systems cost and last.

The bill floor NEM 3.0 doesn’t touch

Even a perfectly sized solar-plus-battery system won’t zero your California bill, and it’s better to know that going in. Two charges survive no matter how much you generate:

• Non-bypassable charges of roughly 2–3¢/kWh on the power you still draw, which you can’t offset with exports — commonly $25–$35/month on their own.
• The state’s new income-graduated fixed charge, phased into SDG&E, SCE, and PG&E between late 2025 and March 2026 at about $24/month for a standard residential customer (less for CARE and FERA customers), in exchange for per-kWh delivery rates dropping around 10%.

Together those set a realistic monthly floor. A good NEM 3.0 system shrinks your bill to that floor and shields you from the worst peak rates — it doesn’t make the utility disappear. Any quote promising a literal $0 bill is overselling.

The incentives that still help

With the federal residential credit gone, California-specific programs matter more:

• SGIP (Self-Generation Incentive Program) is closed as of late August 2026. Every step is fully subscribed and the budget is exhausted — general-market, equity and equity-resiliency alike. Historically the program ran roughly $150 to about $1,000 per kWh depending on tier, with the large numbers reserved for qualifying low-income and medically vulnerable households, but none of that is available to a new applicant today. SGIP has reopened with fresh budget cycles before, so check the program’s own status page rather than assuming it is finished permanently. What you should not do is put an SGIP figure into a 2026 budget or accept a quote that already subtracts one.
• §48E via lease/PPA. A cash or loan buyer gets $0 federal credit in 2026, but a third-party-owned system can still capture the 30% commercial credit — see can you still get 30% in 2026 .
• High retail rates themselves are the biggest “incentive.” Ironically, California’s expensive power is what keeps solar worthwhile even after the export cut.

If you’re already on NEM 2.0

Don’t panic-read this as applying to you. NEM 3.0 only affects systems whose interconnection application came after April 14, 2023. If you’re grandfathered on NEM 2.0, you keep those retail-rate export credits for your grandfathering window — generally around 20 years from interconnection. That grandfathering is also worth protecting: significantly expanding an older system can trigger a move to NEM 3.0, so check before you add panels.

One thing that’s now settled: the race to newly lock in NEM 2.0 is over for PG&E, SCE, and SDG&E customers. Those utilities have been hitting their state-mandated net-metering caps — SDG&E was the first of the three to reach its cap — and the practical deadline to interconnect under the old rules has passed. New rooftop systems in those territories go on NEM 3.0, full stop. If a salesperson in 2026 promises to get you NEM 2.0, treat it as a red flag.

The NEM 3.0 lawsuit: current status (upheld)

Because “will the courts undo this?” comes up in nearly every California solar conversation, here is where the legal fight stands, dated so you can tell if it’s stale. Solar and environmental groups sued the CPUC over the tariff in 2023. In August 2025 the California Supreme Court sent the case back, ordering the Court of Appeal to re-examine it under a less deferential standard. On remand, in March 2026, the appeals court upheld NEM 3.0 in a 3-0 decision , finding the CPUC acted within its authority. And in June 2026 the Supreme Court declined to take the case up again , without comment. Practically: the California net metering changes have been upheld, the legal avenues are exhausted, and NEM 3.0 is the settled rule to plan around — don’t wait on a courtroom reversal that isn’t coming.

Four NEM 3.0 myths a salesperson might tell you

The policy’s complexity is a gift to bad pitches. Four to shut down on the spot:

• “I can still get you NEM 2.0.” Not for PG&E, SCE, or SDG&E customers in 2026 — those utilities have hit their net-metering caps and the window closed. New systems go on NEM 3.0. A promise otherwise is a reason to walk.
• “You’ll have a $0 electric bill.” Non-bypassable charges and the new fixed monthly charge survive any amount of production; a realistic result is a small floor bill, not zero.
• “The battery is basically free after the tax credit.” The 30% residential credit ended December 31, 2025. A cash or loan buyer pays full price for storage in 2026; only a lease/PPA can still route the commercial §48E credit into the deal.
• “Just oversize the panels and sell the extra back.” Under net billing that surplus is credited at a few cents and any year-end excess settles at the low Net Surplus Compensation rate. Oversizing to export is close to giving power away — size for self-use instead.

If a quote leans on any of these, treat its savings math as suspect and ask to see the assumed export rate, TOU schedule, and monthly floor in writing.

So is solar worth it with NEM 3.0?

Usually yes — but the reason changed. It’s no longer “sell power back at retail”; it’s “cover your own brutal evening rates with stored sunshine.” California solar still pays because retail power is among the nation’s priciest, and a well-sized solar-plus-battery system in the big-three territories still models a 7–11 year payback. Size for self-consumption, budget for a battery that spans the 4–9 p.m. peak, and model the real spread rather than trusting a generic savings promise — with SGIP closed and the federal credit at $0, the spread is the entire case. Start with your numbers on the California cost and payback page , see how the policy fits the national picture in net metering explained , and check California’s line in the incentives-by-state comparison .

Sources

• California net billing (NEM 3.0) — California Public Utilities Commission, net energy metering .
• Electricity rates — U.S. Energy Information Administration, average price by state .
• State and utility programs — DSIRE program index , Database of State Incentives for Renewables & Efficiency, NC Clean Energy Technology Center.
• Federal credit status — IRS, Residential Clean Energy Credit ; Public Law 119-21 (2025).

Figures on this page are estimates built from the sources above, not quotes. State and utility programs change and are often budget-limited — verify current terms on DSIRE and with your own utility before relying on them. Nothing here is tax, legal or financial advice.

Frequently asked questions

Sources: U.S. Energy Information Administration (EIA) electricity rates · DSIRE incentive records · public IRS/OBBBA guidance. Figures are modeled estimates, not quotes or tax advice. See our methodology.

Where to go next

• Battery worth-it calculator — NEM 3.0 lives or dies on the peak-rate-minus-export-credit spread. This prices yours instead of quoting an average.
• Are solar batteries worth it? — The national version of the storage question — what a Powerwall-class unit actually costs in 2026, and the degradation nobody quotes.
• Net metering explained — Where California’s net billing sits against full-retail states and the low-export states, so you can see how unusual it is.
• California cost and payback — State-specific prices and rates on one page, without the national averaging.
• The 2026 net-billing shift — Other states are copying this tariff design. This tracks who has switched and what the switch costs a homeowner.