Two things are true about US solar right now, and only one of them is showing up in the headlines. The install-volume line is bending, and the demand-validation line keeps setting records. Both need to be in the same paragraph or the story is wrong.
The 91% share. Solar plus storage accounted for 91% of nameplate generating capacity added to the US grid in Q1 2026, per data compiled from Wood Mackenzie, the American Clean Power Association, and EIA. That is the highest quarterly share on record. Every other resource class combined, gas, nuclear, wind, hydro, took the remaining 9%.
The volume drop. The share number does not mean the sector is booming. SEIA and Wood Mackenzie put Q1 utility-scale solar at 5.9 GWdc, down 34% year over year and 45% quarter over quarter. Total solar installs fell 27% YoY to 7.8 GWdc. That is the OBBBA hangover: the July 4 beginning-of-construction cliff pulled forward every marginal 2026 project into 2025 safe-harboring, and Q1 2026 was the airpocket.
Why the share still matters. In a period when gas queue positions have exploded (natural gas is now 106 GW of PJM’s reopened queue, up from a rounding error two cycles ago), solar plus storage is still the only combination consistently reaching commercial operation at scale. Queue share and interconnection outcome are not the same variable.
The grid-record scoreboard. Six US electricity records fell in H1 2026, and four of them landed in July:
- California’s battery fleet discharged 12.99 GW on July 9, covering 36% of CAISO evening demand.
- CAISO solar hit 23 GW on July 10, supplying 72% of regional load, third record broken since June 1.
- Texas solar reached 35.4 GW on July 9, on grids that had 21.2 GW of installed utility-scale solar and 29.0 GW of storage.
- SPP, ISO-NE, MISO, and PJM each set new solar generation records within a three-month window, several coinciding with peak load records.
- Solar out-generated coal in May for the first time in US history, roughly 13% of national supply.
Thesis read. The demand-vector case (grid firming, evening peak-shaving, load-record coverage) is validating in the physical grid data even as the policy and volume case takes a hit. The gap between Q1’s 91% share and Q1’s collapsed absolute-install number is where investors should be reading, not at the FSLR guide or the LIT price. Watch Q3 install prints once the safe-harbor pull-forward flushes and cluster-cycle awards start converting: if the volume line bounces back into a grid that just spent H1 setting records, the 2027 bookings picture reprices.
Sources: CleanTechnica (July 16, 2026), Renewable Energy Magazine (July 20, 2026), Utility Dive / Wood Mackenzie US Energy Storage Monitor (June 23, 2026), SEIA/WoodMac US Solar Market Insight Q2 2026.