Solar in California

The California Public Utilities Commission's April 2023 decision adopting NEM 3.0 — formally the Net Billing Tariff — was the most consequential residential solar policy change in a generation. It cut export compensation by roughly 75% compared to NEM 2.0, ending the era when a rooftop array functioned as a full-retail-priced virtual battery on the grid. Existing NEM 2.0 customers were grandfathered for 20 years, so a large installed base still operates under the older, more favorable terms, but new interconnections after April 2023 fall under net billing.

The strategic implication is clear: batteries are now central to the value proposition. Under net billing, a kilowatt-hour stored at noon and discharged at 7 PM replaces a peak-rate purchase worth several times what the grid would have paid for the export. SGIP accelerates this by rebating storage capacity, and the program's equity tiers — aimed at low-income households and disadvantaged communities — can push storage rebates to $925/kWh. The state's property tax exclusion for active solar systems (running through 2025, with extension discussions ongoing) rounds out the incentive stack.

Looking ahead, the policy question is whether net billing rates will be revisited as installations slow. The IOUs and solar advocates continue to spar over the tariff's design, and the CPUC has signaled openness to tweaks that could modestly improve export compensation. For now, the planning assumption should be that self-consumption and storage — not export — drive California's solar economics.

California averages about 5.0 peak sun hours per day statewide, but that single number hides enormous regional spread. The Imperial Valley, Coachella, and inland deserts routinely exceed 6.5 peak sun hours, while the fog-prone coast from San Francisco Bay south through Monterey can dip below 4.5. PG&E territory in the Central Valley — Fresno, Bakersfield, Stockton — sits comfortably above the state average and ranks among the most productive solar regions in North America.

Because NEM 3.0 penalizes overproduction that gets exported at a few cents per kWh, optimal sizing in California has changed. Rather than maximizing panel wattage to push the largest possible surplus onto the grid, well-designed systems now target 80–100% of annual consumption, with the array's midday peak matched to air-conditioning and EV charging load. A south-facing tilt at 20–25° still maximizes raw annual kilowatt-hours, but west-facing arrays — which push production into the 4–8 PM peak window when rates are highest — often pencil out better on time-of-use plans.

Smud and LADWP, the two largest municipal utilities, operate outside CPUC jurisdiction and set their own solar programs. SMUD maintains a simpler, flatter rate structure that makes production sizing more straightforward than in investor-owned utility territory.

At $3.80 per watt, California sits well above the $3.07/W national average, making it the most expensive of the ten largest solar markets. The premium reflects several factors stacked together: high labor rates, costly permitting and inspection regimes in many municipalities (though the state's SolarApp+ rollout is slowly streamlining this), and strong demand that keeps installer margins healthy. A typical 8 kW system runs about $30,400 before incentives.

The 30% federal tax credit returns roughly $9,100 on that system, and SGIP can knock thousands more off a paired battery. Combined, these incentives can bring the effective cost of a solar-plus-storage project down by a third or more — which matters, because under NEM 3.0 the storage component is what actually drives the savings. Homeowners in PG&E, SCE, and SDG&E territory who qualify for SGIP equity adders can see storage rebates large enough to nearly halve the battery's price.

Payback stretches to roughly 18 years on a grid-tied-only system at current rates, but that figure compresses meaningfully with storage and disciplined load-shifting. California's high retail rates also mean each self-consumed kilowatt-hour is worth more, so every panel that offsets peak usage pays back faster than the average suggests.