Digging deep for super hot geothermal

The appeal starts with a physical constant rather than a technology. Araque frames the target as temperature, not depth: around 800 degrees Fahrenheit, and no hotter, because that is where water’s thermophysical properties, higher density and lower viscosity, combine with heat-to-electricity conversion efficiency to give the best return. Go above it and you get diminishing returns; below it and you leave value behind. The payoff he claims is disproportionate. The same eight-inch wellbore that moves roughly one to ten megawatts of electric-equivalent energy at 200 degrees Fahrenheit moves about ten times that at 800, so a temperature two to four times higher buys ten times the power. That ratio is the entire economic argument, because it means a well can cost roughly ten times more to drill and still break even. It also changes which side of the cost equation you are working on: rather than driving drilling costs down, you drive output per well up, which is why he expects drilling to fall from about half of a conventional geothermal project’s cost to roughly 20% to 30% of the per-megawatt-hour figure for super-hot systems.

The geographic promise is real but staged, and the episode is careful about this in a way the framing around it is not. Eight hundred degrees exists everywhere if you go deep enough, but the depth ranges from about three miles to about twelve depending on location, and the near-term targets are the Pacific ring of fire, the Atlantic ridge including Iceland, and Kenya. Those are broadly the places that already have geothermal. Araque’s own staging follows: build what he calls tier one, the three-to-five-mile shallow super-hot systems, first, and progress to the deep ones later, because starting deep would be unnecessarily hard. For the shallow tier the materials problem is mostly incremental. Cements that cure at high temperature already exist, elastomers have to be designed out because rubber will flow, steels hold up because power plants already run hotter, and the real limit is electronics, which fail much above 400 degrees Fahrenheit but can be kept cool by circulating fluid. Deeper systems reopen those gaps, and he is explicit that closing them needs an industry with enough momentum to make suppliers innovate.

The turn in the conversation is the most useful thing in it. Kann proposes two technical risks, drilling to depth at temperature and inventing a form of fracturing that works there, and Araque initially ranks them clearly: drilling far outweighs fracturing, because fracturing has precedent in nature wherever hydrothermal vents form, while drilling to those conditions from the surface has no precedent at all. Then, describing the first project, he says the opposite is true of the tier he is actually building. The Oregon site already has holes at the right temperature and depth, drilled in the 1980s or 1990s by a developer hunting for a conventional hydrothermal resource, which it failed to find, and abandoned; Araque counts more than fifty such wells worldwide. Kann catches it and states it back: in shallow super-hot systems the drilling is proven, so the remaining challenge is building the fracture network. Araque agrees. This reverses the intuitive reading of the company, because the exotic drilling technology is what unlocks the deep tier later, not what makes the first project work. And it puts all the near-term risk on the part with the thinnest evidence. Permeability falls as you go deeper under lithostatic pressure, but his argument is that the crust is already critically fractured and that injecting cold, dense fluid into hot rock opens that network through density contrast alone, without the sustained high-pressure surface pumping that oil and gas fracturing requires. Asked directly whether anyone has ever done it at these depths and temperatures, he says no. The world’s ultra-deep holes are cold. The supporting evidence is laboratory work and an inference he drew from oil and gas literature on lost circulation events, where crews lose drilling mud past a certain depth-temperature threshold, which he reads as the same permeability activation happening accidentally.

That makes the flow test the only milestone worth watching, and he says so: two wells connected by a fracture network, producing steam at a stated temperature, pressure and flow rate, holding both without decline. Everything after it, he argues, is a conventional surface power plant. His roadmap puts that first flow test in Oregon by the end of 2026 at roughly three miles, subcritical, at 25 to 30 megawatts of electric equivalent, with a deeper and supercritical version in 2028, while the drilling technology advances separately toward five kilometers in 2027 and ten in 2028. The economics he attaches are $50 to $100 per megawatt-hour, with the shallow systems at the low end and the price holding anywhere in the world. Two things about that number deserve to travel with it. It is a target from a founder describing a system that has not yet produced power. And the benchmark he chose is telling: he wants each well to match the energy output of an oil and gas well, on the theory that anything less will never attract the oil and gas industry to participate at scale.