The Reality of Solar in San Francisco

Solar in San Francisco works. It just doesn’t work the way the sales pitch describes.

The pitch is simple: panels go on the roof, the bill disappears, PG&E pays for your extra power, and the system eventually pays for itself. The real version has fog, fixed charges, weak export credits, and a battery that changes everything.

The same billing fine print showed up on the gas side when Peak Six made my PG&E bill worse. Utility math punishes anyone who stops at the headline rate.

The Anonymized Data

This is from one small San Francisco house with rooftop solar and a home battery. I rounded the numbers and removed account details, exact contract amounts, equipment serials, and bill PDFs.

Category	Rounded value
Solar system size	about 4 kW DC
Battery size	about 13 kWh
Installed cost before incentives	mid-$20k range
Installer production estimate	about 5 MWh/year
Actual data window	about 10 months
Actual solar generated	about 3.3 MWh
Home electricity used	about 4.3 MWh
Solar coverage	about three quarters of home use
PG&E electric bills paid	about $600
Estimated bill without solar	about $1,900
Estimated savings so far	about $1,300
Energy exported to PG&E	about 100 kWh
Export credits received	only a few dollars

That last line is what matters. The system made a lot of useful power and almost no money selling it back.

What The System Actually Did

The solar covered most of the house’s electricity use by keeping the power inside the house first. That’s the part that actually pays well — and under current PG&E billing rules, it’s essentially the only part that does.

Period	Solar generated	Home use	PG&E bill	Estimated no-solar bill	Estimated savings
First operating data window	about 3.3 MWh	about 4.3 MWh	about $600	about $1,900	about $1,300

The savings came from avoiding PG&E purchases, not from selling extra power.

Selling Power Back Was Basically Nothing

The system exported about 100 kWh to PG&E across the whole data window, earning only a few dollars in credits. That’s not a rounding issue — it’s how the current billing rules work.

Exported energy	Export credits
about 100 kWh	only a few dollars

The money is in self-consumption: use the solar in the house, charge the battery, avoid buying expensive evening electricity from PG&E. Selling daytime power back isn’t a real revenue stream anymore.

NEM 2 vs NEM 3

Under older NEM rules, exported solar received credits closer to the retail rate, which made solar-only systems easier to justify. Under the newer Net Billing Tariff (often called NEM 3), exported energy is credited at its grid value at that moment — usually lower than retail, though it can spike during late-summer evenings. PG&E calls this the Solar Billing Plan.

Rule	What matters
NEM 2	Exporting power was more valuable. Solar-only economics were stronger.
NEM 3 / Solar Billing Plan	Exporting midday power is usually less valuable. Batteries matter more.

The battery isn’t just a backup toy under NEM 3. It stores daytime solar and shifts it into the evening when PG&E electricity is expensive. Without it, more daytime power would spill to the grid at weaker export values.

What It Costs To Own

The installed cost was in the mid-$20k range before incentives, and the federal tax credit is a significant part of the math.

The IRS says the Residential Clean Energy Credit was 30% for qualified clean energy property installed through December 31, 2025, after which it’s no longer available. Solar panels and battery storage (3 kWh capacity minimum) both qualify. The credit is nonrefundable but can be carried forward to future tax years.

I’m not putting my exact tax or contract numbers here. In round numbers:

Item	Rounded impact
Installed cost before incentives	mid-$20k range
Federal credit, if eligible	about 30% of qualified cost
Net cost after federal credit	high-teens range
Current annualized savings	roughly low-to-mid $1k/year
Payback before incentives	mid-to-high teens in years
Payback after federal credit, if eligible	low teens in years

Without the tax credit, the payback stretches significantly. With it, the project starts looking more like a normal long-term home improvement.

There are still caveats worth knowing:

Caveat	Why it matters
You need tax liability	A nonrefundable credit only helps if you owe enough tax over time.
Placement date matters	The IRS says the property must be installed by the deadline.
Qualified cost matters	Not every related home project is automatically part of the credit.
Rate changes matter	PG&E rates and export values can change over time.
Battery replacement matters	A battery may not last as long as the panels.

This isn’t tax advice — it’s the ownership reality. The system isn’t just “cost divided by monthly bill savings.” It’s gross cost, tax credit, PG&E rate design, battery behavior, and how long you stay in the house.

The Payback

The current data shows about $1,300 of estimated savings over the first operating data window — roughly low-to-mid $1k per year annualized. I wouldn’t overfit that number. The window is missing a complete summer and a full PG&E annual true-up.

Scenario	Rounded payback
Conservative case	closer to the mid-teens
Better case with tax credit and strong self-consumption	closer to the low teens

It’s a long-lived home improvement that lowers the electric bill, provides battery backup, and slowly pays back over time. The payback is real but not fast.

Why San Francisco Is Weird

San Francisco is not Fresno. The roof gets fog, summer can be gray, and peak production months don’t always land when you’d expect. Winter solar is lower, evening electricity still matters, and PG&E fixed charges still exist. Solar helps, but it doesn’t make the utility disappear.

What Data Is Still Missing

Missing item	Why it matters
Full first year of Tesla daily data	The current data starts after the system was already coming online.
Summer 2026 production	June and July should show peak generation.
Full annual PG&E true-up	This confirms final credits and charges after a full billing cycle.
Long-term degradation	Panels slowly produce less over time.
Maintenance and repair history	Ownership cost isn’t just installation cost.

The big picture is still clear: solar is already reducing the electric bill, the battery matters, and selling power back isn’t the business model.

My Take

Solar in San Francisco is worth considering if you use the power yourself, have a battery, and plan to stay in the house long enough. It’s a weaker case if you expect PG&E to pay much for exported power.

The old sales pitch was “your meter spins backward.” The current reality is “keep as much solar as possible inside the house.” For this house, that’s working. The payback is real but slow.