Two miles beneath a volcano in Oregon, a Texan company run by an oil engineer has reached the hottest rock anyone has ever tried to make electricity from. Now it wants to go a mile deeper, until the water it pumps down stops behaving like water.
Mazama Energy, incubated by Khosla Ventures and launched in 2023, raised 135 million dollars in September from a group led by John Arnold's Centaurus Capital and Doerr Capital, with ConocoPhillips and Shell Ventures joining. The target is rock at 400 °C. At that temperature and pressure water becomes "supercritical", neither liquid nor steam, and carries far more energy to the surface. One such well, the company says, could produce as much power as ten conventional ones.
The problem: geothermal is clean, steady and almost everywhere, and still too expensive
Geothermal is the renewable that does what AI data centres, grids and factories want: firm, carbon-free power, round the clock, from a small footprint. Earlier in this series Atomarine and Starcloud left the land to find power, and ThinkLabs tried to teach the grid to cope. Geothermal promises the power without leaving. The catch is that conventional geothermal needs natural pockets of hot water, which exist in only a few places, and the newer approach, enhanced geothermal, drills miles down and fractures dry rock to make an artificial reservoir. Fervo Energy proved that works, went public this year and signed a large power deal with Google. But at about 200 °C it takes many wells and a lot of water to make a plant, and a rig can cost more than 100,000 dollars a day. "Superhot rock is almost essential for geothermal to become cost-competitive with fossil fuels," says Mazama's chief executive, Sriram Vasantharajan.
What it actually does
The physics first. Heat water past 374 °C under enough pressure and it stops boiling; it becomes a dense fluid that carries far more energy per kilogram than steam. Mazama's estimate is that a well at 400 °C could deliver up to ten times the power of a well at 200 °C, with 75 per cent less water and 80 per cent fewer wells.
The engineering is oil-field engineering pointed at a volcano. On the western flank of Newberry Volcano in central Oregon, which the company calls one of the largest geothermal reservoirs in the United States, Mazama drills a pair of wells. Cold water goes down one. The rock between the two is cracked open using a method the company calls Thermal Lattice, which pulses pressure below the level that would create new fractures in order to wake up the natural ones. The water heats as it crosses the cracks and comes back up the second well to a turbine. The hard part is not the idea but the tools: steel, cement and electronics that survive 400 °C in granite and basalt, and drilling that does not take months.
What it has done so far. A rig arrived in October 2024. By late October 2025 the company reported 629 °F, or 331 °C, at the bottom of a 10,200-foot deviated well, the hottest engineered geothermal system on record, landed within six feet of its planned path, with circulation tests confirming the two wells were connected. Average drilling speed was 76 feet an hour through granite, basalt and granodiorite, with no tool failures, and the work was written up for the Society of Petroleum Engineers' annual conference. In September 2026 a second well, Athena, reached 10,350 feet in about fifteen days, roughly 80 per cent faster than the first; it is heading for about 15,500 feet and rock above 400 °C. No electricity has been generated yet.
Who is behind it
Sriram Vasantharajan, the chief executive, is a chemical engineer who spent his career on difficult oil and gas wells, deep water and heavy oil among them. The company's capabilities, in its own words, come from its founders: AltaRock Energy, a geothermal developer, and Blade Energy Partners, an upstream oil engineering firm. Khosla Ventures incubated it, Gates Frontier backed it early, and in 2024 the US Department of Energy awarded it a 20 million dollar grant for the Newberry demonstration, alongside three national laboratories and two universities. The Series B adds ConocoPhillips and Shell, which is to say two oil majors now own a piece of a geothermal company run by oil people. "Superhot geothermal offers the promise of firm, carbon-free power with a relatively small footprint," said Arnold.
Where it is heading
Project Ceres is next: horizontal wells into superhot rock, backed by the Department of Energy, designed for about 15 megawatts each, with drilling later this year and tests from early 2027. Vasantharajan calls it the blueprint for commercial deployment. The local paper reports a 50-megawatt plant feeding the grid in 2029 and 200 megawatts the year after, and the company's estimate of what the Newberry site could eventually hold has risen from 5 gigawatts to 10. The customers it talks about are central Oregon's utilities and the data centres moving in. The price it is aiming for is under five cents a kilowatt-hour.
An investor's eye
What has to be true: that wells can be drilled, cased and kept open at 400 °C for years, not days; that the cracked rock carries enough flow to run a turbine, which no test has yet shown; and that the economics beat the 200 °C enhanced geothermal that Fervo is already selling. The biggest risk is the one in the timeline. The record was set a year ago and the first electricity is still three years away, and dates in this business move: one report in September put first power at "next year", the local paper at 2029. It is not alone on the mountain either. Quaise Energy raised 180 million dollars in August for a site just south of Newberry's national monument.
It is a bit like the difference between a kettle and a pressure cooker. A kettle boils. Push the pressure and the heat far enough and water stops boiling altogether and becomes something else, with far more energy packed in. Mazama's bet is that the something else is worth drilling three miles for.
To sum up: Mazama has drilled the hottest geothermal well on record and has not yet made a watt. The next hole decides whether superhot rock is a power plant or a science project. Watch early 2027.
Sources
- TechCrunch, Khosla-backed Mazama Energy just raised $135M to drill deeper into superhot-rock geothermal (17 Sep 2026): 15,000 ft and 400 °C; 15 MW per well; the ten-times claim; the site estimate up from 5 GW to 10 GW; first power "sometime next year"; 200 MW by 2030; the full investor list.
- Canary Media, Mazama Energy raises $135M for superhot geothermal at Oregon volcano (17 Sep 2026): Frisco, Texas; launched 2023; the CEO's oil and gas background; the Series B leads and participants; 331 °C; 10,350 ft, about 80% faster; rigs at over $100,000 a day; Project Ceres; the Vasantharajan and Arnold quotes; Fervo, Sage and Quaise.
- Bend Bulletin, Mazama Energy raises $135m for Newberry geothermal project (18 Sep 2026): first electricity to the grid in 2029 from a 50 MW facility, 200 MW in 2030; the Athena well at 10,350 ft after about 15 days; the 15,500 ft target; water's critical point; 75% less water and 80% fewer wells; sales to central Oregon utilities and data centres; Quaise's site south of the monument.
- JPT (Society of Petroleum Engineers), Mazama Energy reports record 629 °F enhanced geothermal system at Oregon volcano (Oct 2025): the injector and producer layout; 10,200 ft; the Thermal Lattice method and SPE paper 228048; 76 ft/hr average and the 2,760 ft bit run; the $20m DOE grant; the under-5-cents target.
- Central Oregon Daily, Oregon project reaches 629°F, redefining geothermal power (28 Oct 2025): landed within six feet of plan; connectivity confirmed; the 15 MW pilot then 200 MW; "one of the largest geothermal reservoirs in the U.S.".
- BusinessWire, Mazama Energy Awarded Grant by U.S. Department of Energy for Superhot Rock Geothermal Demonstration Project (21 Nov 2024): the DOE Geothermal Technologies Office; the national-laboratory and university partners; AltaRock Energy and Blade Energy Partners as founders; feasibility testing from October 2024.