New firm generating capacity added to the grid since 2024, cumulative

Why this row

Data centers drew about 175 TWh in 2024 and the bottom-up pipeline puts them near 1,000 TWh by 2030, inside a published range of 290 to 1,050. Against that load the country added roughly 18 TWh a year of new firm-equivalent supply in 2025, when 150 to 250 TWh a year is needed later in the decade. The J-Curve white paper works the gap backwards: net of everything already announced, the firm-capacity shortfall reaches 40 to 55 GW as early as 2027 and 195 to 240 GW by 2035, and only one modeled path closes it, a nation-scale mobilization of three firm levers, about 100 GW of enhanced geothermal, 63 GW of nuclear and 80 GW of gas and solid-oxide fuel cells, roughly 1,800 TWh a year of new firm output by 2035. On that path the country's energy return on investment dips to a trough in 2032 and recovers to industrial levels by 2035; on the do-nothing path it falls to the level at which Germany began deindustrializing in 2022. The row is the mobilization path cut at the plan's window: 170 GW of new firm capacity by 2032. Every part of it is at or past the historical record. Firm sources have been built at about 5 GW a year; gas turbines are sold out through 2030; nuclear at 6 GW a year is above the Messmer-plan pace; geothermal at 100 GW is three times DOE's own liftoff estimate for 2035. Row 5 counts every gigawatt developers add; this row says how much of it has to be firm, and it is the row the AI Transition Tracker already scores every day.

The path

the J-Curve mobilization path: 48 GW by 2027 and 243 by 2035, linear between; the tracker scores this row daily as TWh/yr against the same path

Each July 4 the row shows its baseline, this path, the latest official value and a status of ahead, on schedule or behind. As an anticipatory row it reports the public decisions that delayed it.