Opinion — California's gas burn fell 18% in five years — and LNG's financing model can't survive a thin weathe — Weekend Edition

California's gas burn fell 18% in five years — and LNG's financing model can't survive a thin weather band

California's gas burn fell 18% in five years — and LNG's financing model can't survive a thin weather band Between January and August 2025, California's gas-fired power plants produced 45.5 billion kilowatt-hours, down 18% from the same stretch of 2020. Over the same five years utility-scale solar nearly doubled, climbing from 22.0 BkWh to 40.3 BkWh. The single-year move was the sharpest part: from 2024 to 2025, gas generation dropped 9.5 BkWh while solar added 5.9 BkWh, both moving about 17% in opposite directions. A chart of those lines has been circulating this week, pulled from the EIA's Today in Energy and passed around with a pointed caption — that LNG importers pay two to three times what a Californian generator pays for the same molecule, and that solar has therefore quietly settled the question of LNG's future. The displacement is real. But for anyone underwriting a new liquefaction train, the damaging part of the California data is not the volume of gas that disappeared. It is the shape of what's left. New LNG export capacity is not financed against spot demand. It is financed against twenty-year, take-or-pay contracts with steady, baseload-shaped offtake — the buyer commits to pay for the cargoes whether it lifts them or not, and that commitment is what lets a developer raise ten to twenty billion dollars for a greenfield terminal. The whole edifice rests on a buyer who can promise to burn a predictable quantity of gas every year for two decades. California is showing what happens to that promise. Gas there is no longer a baseload fuel being slowly eroded; it is becoming the residual that fills whatever solar, storage and hydro leave behind. In the 2021 drought, hydro output collapsed to 10.3 BkWh and gas generation spiked to 63.5 BkWh to cover the gap. In a normal 2025, gas is drifting toward 45.5 BkWh and falling. That is a swing of roughly eighteen billion kilowatt-hours between a bad water year and a good one — an enormous weather-driven band, with the midpoint sinking every year as solar capacity climbs. Spiky demand like that is wonderful for spot volatility. It is lethal for a contract that needs a buyer to commit to a flat annual quantity through 2046. The bulls have a retreat ready, and it is worth taking seriously. Yes, they say, midday gas is finished, but that 63.5 BkWh drought year proves gas is irreplaceable insurance — the dispatchable firming fuel that keeps the lights on when renewables fail. Hold gas to that role and LNG demand has a floor. The trouble is that the floor is being colonized too. Total California electricity demand has risen 8% since 2020, yet gas fell anyway. Solar alone does not close that arithmetic — gas dropped 9.5 BkWh year-on-year while solar rose only 5.9 BkWh. The missing piece, roughly 3.6 BkWh, is batteries and recovered hydro eating the evening ramp, the early-evening hours after the sun drops when gas peakers used to earn their keep. That ramp is precisely the niche the firming-fuel thesis depends on, and it is the niche storage was built to take. Even granting gas its insurance role, insurance throughput is unbankable. A plant that runs flat-out one year in five and idles the rest generates revenue far too lumpy to support a twenty-year take-or-pay obligation. This is the split that matters: the capacity stays, the fuel leaves. California will keep its gas fleet on standby for the next drought, just as SPP's own Market Monitoring Unit this month called resource adequacy "one of the most critical issues" facing its footprint — nobody is retiring firm capacity into a tightening grid. But capacity is not consumption. The megawatts of installed gas turbines can hold flat while the gas actually burned through them falls a fifth in a year. Anyone modelling LNG demand off the installed gas fleet is reading a number that no longer tracks the burn beneath it. The standard rebuttal is that California is a rich, policy-driven outlier and that LNG's growth was always going to come from emerging Asia. The EIA's own write-up concedes as much. But the week's other headline complicates that escape route. The IEA flagged that households across the Philippines and its Southeast Asian neighbours are still being squeezed by elevated energy costs — costs driven in large part by what they pay for imported gas. The price gap is not subtle: Asian LNG settled near $15.31 this week against Henry Hub at $3.20, leaving importers paying close to five times the US benchmark for the same fuel. The countries being sold as LNG's growth engine are tropical, high-insolation, and acutely bill-sensitive — which gives them both the resource and the motive to build solar and walk away from the import meter. LNG's pricing power in its growth markets is the very thing arming its competitor. California is not the bellwether for the United States. It is the preview for Manila. None of this means the molecules stop flowing tomorrow. Global LNG volumes can keep grinding higher for years on Asian demand and coal-switching, and the spot market can stay loud and profitable — TTF at $40.64 and a brutal 18.7% single-day drop on Friday is exactly the kind of churn traders are paid to ride. A booming spot market and a dying project-finance market are not contradictory; they are the same picture seen from two desks. The structure that converts LNG demand into sanctioned terminals — long-dated, flat-shaped, take-or-pay — is the part California is quietly dismantling, one drought-to-normal swing at a time. For a trader, that volatility is the opportunity. For anyone signing off on a greenfield liquefaction FID in 2026, it is the line between a plant that's sold out for two decades and one that spends them looking for a buyer willing to promise something California no longer can.