CAISO has more batteries than the rest of the country combined

Jeff Burka
Singularity
Published in
3 min readApr 17, 2024

--

Today I was glancing at Carbonara, our dashboard for real-time grid carbon insights, and something caught my eye.

A screenshot of the Generated Fuels Trend chart from Carbonara, filtered to show solar and batteries. The batteries line has two significant peaks when solar drops to 0 and when it ramps up again.

That’s real-time data showing over 4 gigawatts of power in California coming from utility-scale battery storage, a resource that was almost nonexistent just a few years ago. Why did it get so big so quickly?

California has a lot of solar power. This is critical for reducing the carbon emissions of the electricity system, but during some parts of the year (like right now), it causes a challenging phenomenon known as the “duck curve”.

Take a look at this graph, showing demand vs. wind and solar generation on March 27th, 2024:

A graph showing demand, net demand, and solar over a day. In the middle of the day solar spikes up to a plateau for ~9 hours, demand dips modestly and net demand dips steeply during this period.
The net demand trend from CAISO’s Today’s Outlook

See the duck? Me neither, but that’s what it’s called. The term “duck curve” was first coined in 2013 by some particularly imaginative analysts at CAISO, the California grid operator, who were staring at this graph:

A chart titled “Figure 2: The duck curve shows steep ramping needs and overgeneration risk”. The chart shows net load over a day (March 31), with a line for each year from 2012 to 2020. 2012 and 2013 are actualy data while 2014–2020 are predictions. As the year increases, the dip in the middle of the day gets ever deeper and the increase around 6pm gets ever steeper.
It looks like a duck, I guess.

Net load is the amount of power being supplied by the grid, not including power generated by wind and solar. This concept is useful to grid operators because wind and solar are out of their control — they just happen when they happen. Net load represents the amount of power that is within their control, and must be supplied by instructing power plants to spin up and down.

The graph above illustrates the challenge: as solar becomes more common, net load during the day will drop lower and lower, until eventually solar might cover all demand. However, load when the sun is down will be as high as ever, and likely increase as more of our society becomes electrified. This causes two problems. First, we might be generating more solar than we need, especially considering existing resources like nuclear power plants that can’t just turn off and on at sunrise and sunset. What do we do with all that extra power? Second, the sun tends to go down at exactly the same time that demand is spiking in the evening. How do we ramp up gigawatts of power fast enough to replace all that disappearing solar?

One answer to both of these questions is batteries. When the sun is blazing and there’s more solar than we need, we can use that excess power to charge up batteries. Then we can call on those batteries at sunset when net load is ramping up fast, essentially converting unpredictable solar power into a stable dispatchable resource under the control of the grid operator.

This is a simple strategy, but CAISO is implementing it at levels way beyond the rest of the country. As of the latest reported data, they have more battery capacity (8,060.5 MW) than every other US balancing authority combined (7,934.6 MW). Look at how fast this resource appeared in just a few years:

Data from EIA’s Monthly Electric Generator Inventory

This is an exciting trend, and it’s only part of the story. These charts don’t include the huge amount of behind-the-meter storage already installed in homes and businesses, nor do they include other storage technologies like pumped hydro. As renewable generation continues to spread, we can expect to see this explosive growth in storage continue, and hopefully see other grids start to catch up with CAISO.

If you’re interested in keeping up with this trend, check out the live data yourself, and look into our grid carbon data API or our open source emissions project to go deeper.

--

--