A startup claims to have uncovered a hidden source of geothermal energy, and the reveal is generating both excitement and debate. Zanskar, a company that uses artificial intelligence to pinpoint subterranean geothermal resources, says it has located a new site in Nevada that could support a commercially viable power plant. The company describes the discovery as the first of its kind in decades for the industry.
This breakthrough builds on years of investigation into how to locate geothermal reservoirs and underscores geothermal energy’s growing potential.
“When we started this company, the prevailing view was that geothermal was a dead end—a record of failed attempts and faded promise,” says Carl Hoiland, Zanskar’s cofounder. “Now, with these new tools and capabilities, you can systematically identify such sites and reduce the financial risk. This feels like the first clear signal that the tide is turning.”
Geothermal power, in theory, is one of the simplest renewable energy methods. It relies on underground reservoirs of hot water heated by the Earth’s core. This steam can drive surface turbines to generate electricity, typically without heavy mining or complex fuel processing. Regions where tectonic plates collide—and thus crust is thinner—are particularly well-suited for geothermal plants. The western United States stands out as a prime candidate. The world’s largest developed geothermal field, located in California, sits on a long history of exploiting hot springs, with the first power plant built there in the early 1920s.
Yet a major challenge lies in locating these resources. Many productive geothermal systems are deep underground with little surface evidence. Known as blind or hidden systems, they require deep exploration to determine whether a power plant is feasible. Consequently, many plants have been built over systems discovered by chance during drilling for agriculture, mineral exploration, or oil and gas activities.
“Finding a geothermal site is like searching for a needle in a haystack,” notes Joel Edwards, Zanskar’s other cofounder. “Only a tiny fraction of land contains a geothermal system.”
During the 1970s oil crisis, the federal government mapped a grid in Nevada to systematically drill for blind systems as part of expanding the U.S. geothermal output. James Faulds, a geology professor at the University of Nevada, Reno and former head of the Nevada Bureau of Mines and Geology, reflects that, given the limited data at the time, the approach made sense even if it now seems less plausible. The government later cut geothermal research funding, channeling money into fracking, nuclear, and solar while industry focus shifted to known systems. Today, geothermal energy accounts for less than 1% of the U.S. energy mix.
Researchers in the field still see vast, largely untapped potential in blind systems across the western U.S. Zanskar contends that its AI-driven method, which organizes and analyzes large geological datasets, can help uncover those hidden resources.
Zanskar’s approach extends the work of scientists like Faulds, who since the 2000s cataloged the characteristics of known systems to develop methods for locating others. In the late 2010s, Faulds led a Department of Energy–funded project to identify Nevada blind systems using data such as fault patterns and electrical conductivity to triangulate likely locations. In 2018, this technique successfully pinpointed a blind system hot enough for electricity, though it wasn’t tested for commercial viability due to its location within a wilderness study area.
Edwards explains that there is substantial overlap between Faulds’s research group and Zanskar; Faulds was Edwards’s master’s thesis advisor and also advised a Zanskar data scientist. “That group demonstrated that blind hot spots could be found at lower costs than in the early exploration era,” Edwards says.
Over the past several years, Zanskar has gathered data across its Nevada sites, with consistent indications of hot spots representing potential blind systems in previously unexplored areas. The company’s milestone this year was confirming that these signals can translate into actual electricity production—an assertion that requires deep underground drilling to verify reservoir temperature. The announcement is framed as proof of concept: the technology can locate viable blind systems. Further testing will still be needed to determine the reservoir’s size and shape and the water flow rate, all critical factors for estimating potential power output.
“There’s a market signal here suggesting that a power connection will be possible someday,” Edwards says.
In recent months, high-profile agreements and media attention have positioned geothermal energy as entering a new era, with much of the enthusiasm centered on enhanced geothermal systems, or EGS. EGS creates geothermal conditions through hydraulic stimulation, reducing the need to find blind systems but introducing water requirements for rock fracturing and some seismic risk. Fervo Energy is a leading name in this space, having signed partnerships with major oil companies, service providers, and utilities, and one of its plants began supplying data center power in 2023.
However, the spotlight on EGS risks overlooking the value of blind-system potential. While EGS uses less water than traditional fracking, it still requires external water for the fracturing process and can induce low-level seismic activity. It also adds complexity compared to simply drilling to a system and building a plant atop it. Hoiland notes that engineering challenges invariably raise costs.
Zanskar’s founders argue that the global energy potential from blind geothermal systems may be far larger than earlier estimates suggested. A 2008 U.S. government assessment estimated that undiscovered geothermal resources could provide a mean power potential of around 30 gigawatts, enough to power tens of millions of homes. Faulds, however, believes the figure may underestimate the real potential by a wide margin, suggesting that tens to hundreds of gigawatts could be within reach from blind systems as drilling technology advances into hotter, deeper conditions. As technology progresses, the practical ability to harness this energy is likely to grow as well.
