Fervo's Utah geothermal plant starts selling power to the grid

In Beaver County, Utah, drills are still turning at Cape Station. The plant's first unit, 33 megawatts (MW), has been switched on, and it is now selling electricity to the grid. Fervo Energy, a Houston-based startup that went public in May, says its project is the first utility-scale next-generation geothermal plant to deliver power in the world. Commercial operations start Thursday, when Shell Energy, a retail power provider in California, begins buying the output.

Conventional geothermal depends on where nature has put hot water and steam underground. Fervo instead drills into hot, dry rock, using techniques borrowed from the oil and gas industry, and creates a reservoir there that produces steam to drive turbines. Geothermal is a small part of the U.S. grid today, supplying 0.4 percent of electricity generation each year, so the question has always been whether the approach could be scaled up. Fervo's earlier 3.5 MW pilot plant in Nevada, online since late 2023, showed that the technology worked. Cape Station, a $2 billion project planned at 500 MW, is the test of whether it works at the size that matters to a utility.

The schedule is precise. Fervo says it is commissioning two more 33 MW power blocks by January 1, 2027, and the remaining 400 MW, supported by a new agreement with Google, is under construction for 2028. Cost is the harder part. Phase 1 is set to deliver power at $7,000 per kilowatt and phase 2 at $5,500 per kilowatt, according to what the company previously told Canary Media. For comparison, building a natural gas plant cost an average of $2,157 per kilowatt in 2025. Fervo has not disclosed the terms of its power purchase agreements, the long-term contracts that sell its output, which together cover about 1,000 MW, enough for roughly 1 million U.S. homes each year.

For the grid, the gain is flexibility as much as clean power. Geothermal has traditionally run around the clock, but next-generation plants could ramp up in the evening, when solar output falls, which could lower overall grid costs. They could also be sited to serve data centers directly, "behind the meter", so that the cost of new capacity is not passed on to utility customers. Those benefits depend on drilling costs falling and on permitting and community concerns being handled, and experts say both remain real barriers. For now, the answer to "what changes" is a first 33 MW unit feeding Californian customers this week, and a path to 500 MW by 2028 if the drilling and the money hold.