The ground beneath our feet holds a lot of energy, so experts at the Department of Energy are thinking about geothermal power can produce 60 gigawatts – or nearly 10% of US electricity – by 2050.
Zanskar co-founder and CEO Carl Holland thinks that the lofty number is too low, mostly because it discounts the potential of conventional geothermal energy.
The DOE figures take into account the progress of enhanced geothermal energy, which uses fracking techniques to access hot rock deep underground. Company like Fervo and Sage Geosystems have been working on that angle, and experts agree that it has tremendous potential. In contrast, conventional geothermal, which emits hot spots of natural cracks, is stagnant, producing only 4 gigawatts in the United States, up only about gigawatts in the last decade or so.
Conventional geothermal has been held back by outdated assumptions, Holland told TechCrunch.
“They are underestimating a number of systems that have yet to be discovered, perhaps by an order of magnitude or more,” Holland said. With modern drilling techniques, “you can get more out of each one, maybe an order of magnitude or more out of each one. Suddenly the number goes from tens of gigawatts to potentially terawatt-scale opportunities.”
Zanskar relies on AI to help relieve conventional geothermal of its pain. In the process, the startup has resuscitated flagging power plants in New Mexico and found two new sites with over 100 megawatts of combined potential.
The success helped Zanskar get $115 million Series C led by Spring Lane Capital with participation from All Aboard Fund, Carica Sustainable Investments, Clearvision Ventures, Cross Creek, GVP Climate, Imperative Ventures, Lowercarbon Capital, Munich Re Ventures, Obvious Ventures, Orion Industrial Ventures, Safarna Group Sustainable Partners, Safarna Group Sustainable. Tranquillion, Union Square Ventures, University Growth Fund, and UP.Partners.
Techcrunch event
San Francisco
|
13-15 October 2026
Many potential geothermal sites have been overlooked because people have looked for evidence on the surface like hot springs or volcanoes, Holland said. But about 95% of all geothermal systems lack that, he added. “We keep discovering by accident. This is a great application for AI.
To find new geothermal sources, Zanskar first feeds a machine learning model that is supervised by some data, including accidental discoveries. When they have identified a promising site, the company sends a team to the site to validate the discovery.
Zanskar then had to come up with a plan to expand. For that, the team used another AI approach, developed for this purpose called Bayesian evidential learning (BEL). In BEL, existing data helps build a number of assumptions known as priors, and then the model can manipulate these hypotheses, generating probabilities for each. Where there are gaps, startups have created geothermal simulators to fill the void.
So far, Zanskar’s approach has worked. A previous round of funding allowed the startup to explore three sites, each of which was deemed a success. “Three out of three,” said Zanskar CTO Joel Edwards. “What do you see when you try 10?”
The company has enough sites in the pipeline to support at least a gigawatt of generating capacity, Holland said. For now, they have focused on the US West, where it has the most potential. He wants to find at least 10 confirmed sites to help the financial investors of the Zanskar court project, which gives him access to lower cost capital than VCs. If the company can do the work, then it can be done the valley of death which has claimed many other climate technology startups.
Holland is not convinced that Zanskar has solved all the challenges associated with finding geothermal resources, but he is confident that the company is on the right track.
“We now know this is the future of exploration,” he said. “It’s going to change the heat of the earth very quickly.”