Opinion — Solar and wind are both significantly cheaper and more effective than any other form of energy — Weekend Edition

Renewables won the cost war. The grid's scarce asset is now firming. Solar and wind are the cheapest electrons ever built. The bill for using them has moved to storage and dispatchable backup, and that is where the price tension now sits. On Friday, day-ahead power settled below zero in two Australian states: minus $11.65 in South Australia and minus $11.46 in Victoria, each down more than 40% on the day. Queensland fell 50% to $39.89, New South Wales 47% to $42.54. Those are not distressed grids. They are grids with so much midday solar that, for stretches of the afternoon, an extra megawatt-hour is worth less than nothing. Hold that against the headline number everyone has been quoting this week. Levelised cost of energy for utility-scale solar now runs $0.03 to $0.06 per kWh; onshore wind, $0.02 to $0.05. Both undercut new coal and new gas outright. The world added 692 GW of renewable capacity in 2025, roughly three-quarters of it solar. In the United States, wind and solar reached 17% of generation, up from under 1% in 2005. The claim that renewables are the cheapest form of new power is not in dispute. It is settled. The trouble is what that LCOE figure measures. The Economist, in the same piece that celebrated a 90% cost decline since 2010, flagged that the number "does not yet account for the system costs that intermittency imposes." The cost of intermittency did not disappear with the panel price. It moved — from the generator to the firming layer — and Friday's negative prints are the receipt. Look at where firm power still commands a premium. Spanish day-ahead settled at $33.14, France at $25.78, among the cheapest in Europe, because both lean on resources that show up when called: Spanish solar at midday, French nuclear around the clock. Now look at the curve. Spanish baseload for Cal+1 is priced at $61.18, nearly double the day-ahead spot. The market is not paying up for electrons. It is paying up for electrons delivered at 7pm in January, when the sun is down and the battery either exists or it doesn't. The contrast the bulls keep citing makes the point against them. New Zealand runs at 94% renewable; Spain sits above 40%. New Zealand manages it because hydro — dispatchable, storable, dammed water — firms the variable supply at no extra capital cost. Spain has no comparable buffer, which is why the same analysts who praise its cheap generation concede it needs far more batteries. Same technology, same LCOE, a completely different system bill. The variable resource is cheap in both. The firming is free in one and expensive in the other. That is where the marginal grid dollar is migrating. IRENA's own benchmark for round-the-clock renewables — solar paired with enough storage to run 24/7 — lands at $54 to $82 per MWh. That is still competitive with new coal at $70-85 and new gas near $100. But notice what is inside the number: the storage, not the panel, is the part that moved it up from three cents. The cheap half is the generation. The expensive half is making it firm. Even the buildout that looks like a pure cost story shows the same pattern. SunZia, the largest wind farm in the United States, came online this month in New Mexico — 3,650 MW across 916 turbines, more than three times the next-largest American wind project. It took almost two decades of permitting and a dedicated high-voltage line to move the power to Arizona and California. The turbine was never the bottleneck. Transmission and timing were. Offshore wind costs, per Carbon Brief, have risen 40% since 2022 even as panels kept falling, because the expensive parts of the system — the steel, the grid connection, the firming — do not follow the module learning curve. The counter-argument is that storage is on the same downslope as solar and will simply mop this up. IRENA projects renewable costs falling another 40% by 2035; battery packs keep getting cheaper; the firming premium compresses on its own. That is plausible and probably right over a decade. But it is a forecast, and the forward curve is a price. Right now, today, Spanish Cal+1 at $61 against a $33 spot, German day-ahead at $106.93 against French at $25.78, two Australian states paying generators to stop — these are the market pricing firm, on-demand power as the scarce thing and cheap midday electrons as the abundant thing. Scarcity is where pricing power lives. So the trade is not long cheap renewables. Everyone is already long cheap renewables; 692 GW a year says so. The position with edge is long the firming stack — grid-scale storage, the hydro and nuclear and flexible gas that hold value when the wind drops, the transmission that moves surplus to where it is short. Solar and wind have won the argument they were having in 2015. The argument that decides power prices for the next decade is being had one layer up, and the cost the LCOE headline left out is the cost that now sets the bill.