CAISO’s solar gains did not make imports irrelevant

The apparent contradiction is the point. In the first five months of 2026, demand in the California Independent System Operator (CAISO) rose 7% compared with the same period in 2024, while net generation fell 19%. Over that comparison, battery discharge tripled and electricity imports from nearby systems doubled, according to the U.S. Energy Information Administration (EIA). EIA also reports that solar generation increased. These observations can coexist because growth in particular sources of in-area supply does not mean that total in-area generation must rise when imports are also part of the supply picture. That is an interpretation of the combined observations, not a claim that imports alone caused generation to fall. (EIA, L015–L016)

Two comparisons, two different questions

EIA reports that CAISO utility-scale solar generation rose 21% and natural gas generation fell 60% in the first five months of 2026 relative to the first five months of 2024. Solar exceeded gas in daily generation on 82% of days in that 2026 period, versus the 21% figure EIA reports for 2024 and 2025. This establishes a substantial change in the solar–gas comparison. It does not establish how much of CAISO’s demand was met without imports: a count of days on which one source outran another is not an accounting of all electricity supplied. (EIA, L006–L008)

Capacity answers another question again. Between April 2024 and April 2026, EIA reports that utility-scale solar capacity grew 19% to 25 GW, net battery-storage capacity grew 79% to 16 GW, natural-gas capacity remained nearly unchanged at 29 GW, and total net capacity grew 14%, or 11 GW. Those are capacity observations over an April-to-April interval, not generation totals for the first five months of either year. Reading the capacity changes as proof that CAISO had less need for electricity from elsewhere would skip the import evidence. (EIA, L009–L011)

What imports change in the reading

EIA describes batteries charging when midday solar exceeds grid needs and contributing power in the evening and early morning. It also reports the tripling of battery discharge alongside higher demand, lower net generation and doubled imports in the five-month comparison. EIA attributes the increase in imports to relatively inexpensive generation becoming available for import, citing more hydropower imports from the Pacific Northwest as drought subsided and imports from a new New Mexico wind project beginning in April 2026. (EIA, L015–L016)

The useful distinction is time-shifting versus sourcing. Storage discharge helps supply power at a different time from charging; an import brings electricity from outside the system. Neither observation, on its own, tells us that the other became unnecessary. Taken together, EIA’s figures describe a system in which solar and storage grew in importance while external supply took a larger role in the reported comparison. This changes how to assess the buildout: solar’s lead over gas is a meaningful generation result, but it is not a measure of local self-sufficiency. (EIA, L006–L008; EIA, L015–L016)

There is also a limit to the explanation. EIA reports the concurrent decline in net generation and doubling of imports, and gives reasons for increased imports; the supplied excerpt does not quantify how much any one change contributed to the generation decline. Nor does it provide an hour-by-hour match between imports, solar output, charging, discharge and demand. It therefore cannot establish whether a particular import displaced gas, covered an evening gap, reduced costs overall or improved reliability. The defensible conclusion is narrower: evaluate CAISO’s changing supply as an integration of in-area generation, storage discharge and imports, rather than treating a solar-versus-gas result—or installed capacity—as a complete demand-supply balance. (EIA, L006–L016)