Two Cannibals Are Eating California’s Duck Curve

What 10 GW of batteries means for wholesale prices, curtailment, and DER economics

A few years ago, the Breakthrough Institute published an influential analysis called “Quantifying Solar Value Deflation in California.” The core finding: solar’s wholesale value had fallen 37% since 2014. Every new megawatt of solar capacity depresses midday electricity prices, eroding the economics for all the solar that came before it. They called it value deflation. The industry shorthand is simpler: solar eats solar.

We’ve been saying for a while now that the same thing is happening with batteries. And the early 2026 CAISO data makes it hard to argue otherwise.

What the data shows

We pulled 3,300 hours of NP-15 day-ahead pricing from CAISO OASIS and curtailment volumes from the CAISO Daily Renewable Reports covering January 1 through March 10, then compared 2025 to 2026 across the same window.

The headline numbers:

January was particularly striking. Curtailment dropped from 130 GWh to 27 GWh — an 80% reduction year over year. February fell 52%. These aren’t rounding errors. California’s grid absorbed dramatically more renewable energy in early 2026 than it did a year prior.

The reason is straightforward: California added roughly 10 GW of grid-scale battery capacity between 2022 and 2025. That storage fleet charges during midday solar surplus and discharges into the evening peak, doing exactly what it was designed to do.

Solar eats solar. Batteries eat batteries.

Here’s where it gets interesting for anyone making capital decisions around distributed energy.

In a wholesale electricity market, prices are set hourly by the most expensive generator still needed to meet demand. As solar and batteries displace gas plants from more hours of the day, prices fall. But the pattern of where they fall tells you something important about what’s cannibalizing what.

Midday prices (when solar generates) dropped $8/MWh year over year. That’s solar eating solar — more panels flooding the same sunny hours with cheap power.

Evening prices (when batteries discharge) dropped $19/MWh. That’s batteries eating batteries — more storage flooding the same peak hours with power they bought cheap at midday.

The evening collapse is 2.3 times larger than the midday drop, in dollar terms. Batteries are cannibalizing their own revenue faster than solar is cannibalizing its own. And the spread between midday and evening — the price differential that is the entire economic basis for grid-scale storage — shrank 38% in a single year.

What this means for building energy managers

If this feels abstract, consider how wholesale price shapes flow through to the rates your buildings actually pay.

TOU rates. The peak vs. off-peak price gradient on your utility tariff is built on wholesale spreads like these. PG&E, SCE, and SDG&E don’t update TOU differentials in real time, but they do adjust them in general rate cases. As the underlying wholesale spread compresses, expect utilities to narrow TOU differentials over time. The economic reward for shifting load from peak to off-peak hours shrinks with it.

Direct access and hourly pricing. If you’re buying power through an ESP or on a real-time pricing product, you’re already living this. The arbitrage between a 1pm kWh and a 7pm kWh is a third smaller than it was twelve months ago.

DER investment decisions. This is where the implications compound. We’ve already seen this movie play out once. NEM 3.0 slashed rooftop solar export credits precisely because midday wholesale prices had cratered — solar ate solar until the regulators repriced the tariff.

The same dynamic is now pointing at every behind-the-meter investment that depends on a steep TOU differential:

🔋 BTM batteries that arbitrage peak vs. off-peak
⚡ Managed EV charging programs timed to avoid peak rates
🏢 Thermal storage sized for off-peak rates

These technologies work. The engineering is sound. But their financial returns are a bet on the persistence of the peak-to-off-peak spread. If grid-scale batteries keep compressing that spread at 38% per year, the payback math on a 10-year BTM battery investment looks different in year 3 than it did at signing.

The broader picture

Zoom out, and the full-year 2024 vs. 2025 CAISO curtailment data tells a consistent story. Solar production grew 12.5% (50,882 GWh to 57,245 GWh). Total curtailment rose 10% in absolute terms (3,423 GWh to 3,766 GWh). But the curtailment rate held flat at 4.5% — meaning batteries absorbed enough surplus to keep pace with solar’s growth. That’s real progress.

It’s also worth noting what didn’t improve. System-level curtailments — caused by grid-wide oversupply rather than local transmission constraints — more than doubled, from 271 GWh to 622 GWh. Local curtailments held flat. The bottleneck is shifting from “not enough storage” to “not enough transmission,” a problem that batteries alone can’t solve.

Where this is heading

Solar ate solar. Now batteries are eating batteries. The energy transition is working in California, and the data is unambiguous. But success is creating a new kind of risk for building operators: the economics you underwrote last year may not be the economics you live with next year.

The operators who come out ahead will be the ones tracking wholesale price trends, questioning vendor proformas that assume static TOU spreads, and stress-testing DER investments against a world where the duck curve keeps getting flatter.

The duck curve isn’t dead yet. But 10 GW of batteries are doing their best.