The world's largest next-generation geothermal plant is about to put its first power on the grid

Fervo Energy's Cape Station in Utah is on track to send its first power to the grid in October, the first commercial-scale enhanced geothermal project to get there. The technology fracks hot dry rock to make geothermal work far beyond volcanic zones, and Google has just signed for 396 megawatts of it. Here is what is real, what is still a target, and why data centres are watching.

The world's largest next-generation geothermal plant is about to put its first power on the grid
TL;DR

Fervo Energy's Cape Station in Beaver County, Utah, is on track to deliver first power to the grid in October 2026, an initial roughly 33-megawatt block, which would make it the first commercial-scale enhanced geothermal project to reach that milestone. The full roughly 100-megawatt first phase is expected in early 2027. Enhanced geothermal (EGS) fracks hot dry rock and circulates water through the engineered reservoir, using oil-and-gas drilling techniques to make geothermal viable far beyond volcanic zones. The full site is being built to 500 megawatts, with permitting headroom towards roughly 2 gigawatts. October is a target, not an achieved event, and "first power" is that initial roughly 33-megawatt block, not the full first phase, let alone the whole plant. Google, chasing clean power for data centres, just signed for 396 megawatts.

Geothermal has spent decades as the energy source that is always about ten years away: clean, constant, and stuck, because ordinary geothermal plants need a rare natural combination of heat, water and permeable rock. A project in the Utah desert is about to turn that perennial promise into a date on the calendar. Fervo Energy's Cape Station is on track to send its first power to the grid in October, and it does it by manufacturing the underground conditions that nature usually withholds. Here is what is genuinely happening, and where the marketing runs ahead of the meter.

What is Cape Station?

Cape Station is a next-generation geothermal power plant under construction near Milford, in Beaver County, Utah, where Fervo broke ground in 2023. It is being built to 500 megawatts in total, and the company holds permitting headroom to expand the site towards roughly 2 gigawatts over time. The headline near-term event is first power: an initial roughly 33-megawatt block, the first of three that make up the roughly 100-megawatt first phase, targeted to start reaching the grid in October 2026.

Two precise points matter here. First, "first power" in October is that initial roughly 33-megawatt block, not even the full 100-megawatt first phase, let alone the 500-megawatt plant or the 2-gigawatt site ceiling, which is permitting potential rather than steel in the ground. Second, October is a target, described as on track as of early September, not something that has already happened. Fervo has said the full 100-megawatt phase should be running by early 2027, so first power precedes full phase capacity.

What is enhanced geothermal, and how is it different?

Conventional geothermal taps a naturally occurring underground reservoir of hot water, which only exists in the right volcanic or tectonic zones. Fervo instead builds its own reservoir: it drills into hot, dry rock and fractures it, then circulates water through the cracks to carry heat to the surface. To reach the deep, hot rock it borrows the toolkit of the oil and gas industry, horizontal drilling and fibre-optic sensing, with wells running laterals thousands of feet long at depths reaching around 15,000 feet.

The significance is geographic. Because EGS manufactures the permeability rather than finding it, it can in principle be sited far more widely, wherever hot rock is drillable. That is why Cape Station is treated as a bellwether for the whole category, not just another plant.

Is the technology actually proven?

Partly, and this is where honesty matters. Fervo's core evidence is a smaller pilot, Project Red in Nevada, a 3.5-megawatt system that has run for more than 600 days with 98.4% uptime and stable reservoir temperatures, which the company published as proof that engineered geothermal wells hold up over time. That is a genuinely strong result, but it is at 3.5 megawatts. Cape Station is the first attempt to prove the same approach at hundreds of megawatts, so it is best read as a first-of-its-kind bet being tested at scale, not a settled outcome. Fervo itself warns in its risk disclosures that it expects continued net losses for the foreseeable future, and its long-run cost goal of around $3,000 per kilowatt is a target, not today's figure.

Why is this a big deal?

Because a dated, commercial-scale first-power event turns geothermal from a theoretical abundance into a real data point, and because the timing meets a genuine market need. The demand is firm, round-the-clock, carbon-free power for the AI data-centre build-out, and buyers are signing up years in advance. On 1 September 2026, Google signed a 396-megawatt power purchase agreement with Fervo, described as the largest EGS deal to date, on top of an earlier 115-megawatt arrangement. It sits alongside a 320-megawatt agreement with Southern California Edison and a 31-megawatt deal with Shell Energy. Fervo, which became the first EGS company to go public in a roughly $1.9 billion IPO in May 2026, now holds around 862 megawatts of firm contracted power across those deals. For more clean-firm-power context, see our look at the US nuclear reactor revival and the future section.

The one caveat worth repeating: "world's largest" of what?

Cape Station is routinely called the world's largest geothermal project, and that needs a qualifier. It is the largest next-generation, enhanced geothermal project. Conventional geothermal fields are bigger: The Geysers in California runs at roughly 1.2 to 1.5 gigawatts, well above Cape Station's 500-megawatt build, let alone its 100-megawatt first phase. The record Cape Station is chasing is for the engineered-reservoir kind of geothermal, which is the new and hard part, not for geothermal overall.

Cape Station at a glance

Detail
LocationBeaver County, near Milford, Utah
First power~33 MW block targeted October 2026; full ~100 MW phase by early 2027
Full build500 MW total (~100 MW near-term + ~400 MW by 2028)
Site potentialUp to ~2 GW (permitting headroom, not under construction)
TechnologyEnhanced geothermal (fracked hot dry rock, horizontal drilling)
Track recordProject Red pilot: 3.5 MW, 600+ days, 98.4% uptime
BuyersGoogle (396 MW + earlier 115 MW), Southern California Edison (320 MW), Shell (31 MW)