Today less than 1% of global (and American) energy comes from geothermal sources. But researchers at Princeton University predict that technical innovations could, by 2050, produce nearly triple the current output of America's nuclear power plants (which supply roughly 20% of the country's electricity). The International Energy Agency reckons cumulative investment in geothermal globally could reach $1trn by 2035, a big jump from the $1bn to $2bn invested in 2024.
Geothermal emits virtually no greenhouse gases during operation and, because Earth's deep rocks are hot all the time, it provides reliable electricity around the clock—unlike wind and solar. It can also provide clean heat and serve as grid-scale energy storage.
Conventional geothermal works in the relatively few locations where temperatures of 150°C to 200°C and permeable fractures occur within 4km of the surface. Firms drill vertically and use the rising steam to turn turbines.
EGS relies on tapping hot impermeable rock, which is much more common than the confluence of permeable fractures and heat needed for conventional geothermal. It typically reaches depths of less than 4km and temperatures of 150°C to 200°C. EGS borrows hydraulic fracturing ("fracking") and multilateral drilling technology developed in the early 2000s by the shale-oil industry.
In an EGS project, engineers first drill a deep well vertically, then rotate the bit and move it horizontally. A second parallel well is drilled some distance away. The two wells do not touch; instead, fractures are created in the rock between them to form an artificial reservoir. Water is pumped down the first well, travels through the fractures, gets heated, and returns through the second well. The hot water warms another fluid that turns a turbine to produce electricity.
The leading EGS firm is Fervo, backed by Google and valued at about $1.4bn. SAGE Geosystems, based in San Antonio, Texas, has also demonstrated EGS production and is pursuing energy storage using similar technology.
In CLS, engineers circulate a working fluid inside enclosed semi-circular pipe systems. The fluid flows down one side, gets heated at depth and returns via the other. A plus is that this works in arid regions, but because it needs more piping and drilling, CLS is more complex and costly. Firms are making progress with CLS in regions where EGS is not an option because fracking is banned or water is scarce.
Canada's Eavor has drilled two vertical wells 4.5km to 5km deep in Germany and linked them with a dozen horizontal wells, each 3km long, to create its underground "radiator". It plans to produce over 8MW of electricity and 64MW of district heating for nearby villages within a few years.
Beyond 8km deep, where the pressure is more than 200 times that at Earth's surface, water enters a supercritical state (neither liquid nor gas) if the temperature is also above 374°C. Supercritical water penetrates fractures easily and yields five to ten times as much energy per well compared with wells using normal hot water. Modelling by the Clean Air Task Force suggests that 13% of North America's land has superhot potential below 12.5km, and tapping a mere 1% could provide 7.5 terawatts of energy capacity.
Previous attempts to harness superhot rock in Iceland, where supercritical fluids lurk just 2km to 3km underground, ran into difficulties. High temperatures and pressures, as well as corrosive chemicals, damage well casings and drilling tools.
Quaise, a Texan firm, has developed a millimetre-wave energy beam (akin to a laser) that can penetrate the hardest rock. It recently drilled a 118-metre-deep hole into granite at up to five metres per hour, far faster than the 0.1 metres per hour that oil-industry kit is expected to manage at superhot temperatures. Mazama, a Texas-based startup, completed a pilot project in Oregon, drilling wells and stimulating fractures through difficult rock at a record temperature of 330°C and 3km deep, with no breakage of equipment.
Geothermal technology can also be used for grid-scale energy storage. SAGE Geosystems has built one of the world's most intriguing batteries in Christine, Texas, demonstrating the ability to store and release 3MW of power to the Texas grid. The firm drills about 3km deep and fractures the rock to create an underground reservoir. Water is pumped from the surface and stored under high pressure underground. When power is required, the well is opened and the pressurised water—forced up by the rock's natural inclination to close the fracture—turns a surface turbine. This "lung" can store power for much longer than lithium batteries. Because the main hardware costs are fixed, unlike with additive battery blocks, the cost per unit of stored energy goes down the longer the system is designed to run.
The intelligence of any discussion diminishes with the square of the number of participants.