Summary: A Nevada geothermal site that Zanskar says it found using artificial intelligence may mark a turning point for geothermal power. The company claims this is the first deliberate, modern discovery of a commercially viable hidden or blind system in decades. If that claim holds up, it matters for developers, utilities, investors, and policy makers because it would show you can systematically find and derisk blind systems instead of stumbling across them by accident.
Interrupt & Engage: Why this announcement breaks the script
Geothermal has been the quiet cousin of wind and solar: reliable, baseload-friendly, but stuck behind a wall—location risk. Zanskar says it found a commercially viable site where there was no surface sign. Hidden or blind systems have been a needle-and-haystack problem. Needle-and-haystack—repeat that. What if the haystack can be mapped?
What Zanskar claims to have done
Zanskar used AI to process large geological datasets and identify a deep reservoir hot enough to generate electricity. Their cofounders say the find validates years of research and a new method: combine historical geological knowledge with modern data science to predict blind systems. The company drilled and gathered enough evidence to announce commercial viability; further tests will map reservoir size, shape, and flow.
Hidden or blind systems explained
Most geothermal power comes from places with visible surface signs: hot springs, fumaroles, faults. But many productive systems are blind—no surface expression at all. Blind systems exist deep, masked by rock and sediment. Finding them by chance has happened during oil, gas, or water drilling. The problem is that most land has no geothermal system. Joel Edwards called it a needle-and-haystack problem. I repeat: needle-and-haystack. That phrase frames the risk: very low hit rates unless you can narrow the search.
A quick history: why geothermal receded
In the 1970s the U.S. government mapped and drilled Nevada to chase blind systems. The data were limited and costs high, so interest waned when other technologies gained funding—fracking, nuclear, solar, wind. Without steady investment, exploration focused on known fields. Since then, geothermal has supplied under one percent of U.S. electricity. That’s a policy and market failure, not a geological one.
How Zanskar’s method builds on past science
This work didn’t start in a vacuum. Geologists like James Faulds cataloged the geological fingerprints of known systems—fault patterns, structural traps, electrical conductivity, heat flow. Faulds’s research and a DOE-funded effort in the late 2010s showed blind systems could be located at lower cost than in the 1970s. Zanskar layers modern machine learning on top of that research, using pattern recognition at scale. The outcome: prioritized targets where the probability of success is meaningfully higher.
Proof by drilling—and what “commercially viable” means
Zanskar’s announcement follows drilling that shows temperatures high enough for electricity generation. That’s a big step. But “commercially viable” is a conditional claim. We still need reservoir geometry, sustainable flow rates, permeability data, and long-term testing. No, this doesn’t mean every AI prediction will pan out. It does mean there’s a validated workflow: predict → drill → confirm. That workflow changes how capital allocates exploration dollars.
Why the market should notice
Edwards said, “There’s a signal to the market with this announcement that there will be a power plug someday.” Mirroring that phrase: a power plug someday. How will utilities, fund managers, and landowners react when a reliable prediction method exists? Will they commit earlier to leases, permits, and off-take agreements? If investors can move from speculative wells to probability-weighted portfolios, financing becomes cheaper and faster.
EGS versus blind systems: two paths, different trade-offs
Enhanced geothermal systems (EGS) create permeability by injecting fluid and inducing fractures—think fracking for heat. EGS is attractive because you do not need a blind system; you make one. But EGS needs injection water, brings low-level seismic risk, and adds engineering complexity and cost. Zanskar’s path is simpler in concept: find hot water, drill, install. Simpler doesn’t mean risk-free. It can reduce some engineering steps and related costs if a natural reservoir has sufficient permeability and flow.
Scale: how big could blind systems be?
A 2008 government estimate put mean undiscovered power at about 30 GW for blind systems. Faulds argues that is conservative and that tens to hundreds of gigawatts are plausible in the U.S. alone. Combine better subsurface imaging, deeper drilling tech, and systematic targeting, and the resource base grows. The physics is straightforward: heat is abundant; the challenge has always been access. Better targeting increases accessible heat.
Risks, unknowns, and where the story could fail
Don’t read the headline and skip questions. Where are the water rights? What will permitting timelines look like? How large and well-connected is the reservoir—single-well flow can’t power a utility-scale plant. What about land use and community consent, especially near wilderness study areas? And what if predictive false positives keep drilling costs up? Saying “no” to complacency: the tech is promising, not a guarantee. How will the team respond when a high-confidence target underperforms?
Policy and public funding: lessons from history
Public funding in the 1970s and later DOE grants seeded important research. That investment paid scientific dividends but was not sustained. If policymakers want a diversified renewables portfolio, targeted funding for blind-system mapping and demonstration projects makes sense. The private sector can scale after initial de-risking. Ask policy makers: what level of early-stage support will bridge discovery and commercial rollout?
Questions developers and investors should ask right now
What due diligence did Zanskar run on the target? What were the false positive and false negative rates in their models? What is the estimated capital cost per megawatt for buildout here, including wells and surface plant? How long will sustained testing take before an offtake contract is credible? These are open questions. How would you allocate capital if the hit rate rises from historical single-digit to a reproducible double-digit percentage?
How to derisk blind-system projects
Layered approach: (1) Predictive modeling to prioritize targets, (2) shallow geophysics and reconnaissance drilling to confirm subsurface structure, (3) short-duration flow tests to estimate permeability and thermal drawdown, (4) staged financing where early capital covers testing and later rounds cover plant construction. Use joint ventures with utilities or green hydrogen offtakers to secure demand early. Commit to incremental investment steps tied to measurable subsurface milestones—consistency breeds investor confidence.
Economic implications and value chain effects
If blind systems can be found at scale, this will shift where the value sits: more value in exploration datasets and predictive tech, less in accidental finds. Oilfield service companies could repurpose skills to drill deep, hot wells. Utilities win a more diverse baseload mix. Communities near promising fields gain jobs but also negotiate land access and environmental protections. Who captures the upside—startup teams, landowners, or utilities—will shape adoption speed.
Social acceptance and community questions
Communities deserve clear answers: noise, traffic, water usage, seismic risk, and long-term land restoration. Confirmation bias can make locals skeptical after past promises. Blair Warren’s persuasion approach matters here: encourage the dream of local jobs and lower emissions, justify past industry failures with honest context, allay fears with transparent testing plans, confirm valid suspicions about risk, and empathize with local concerns. Will project teams commit to open data and third-party monitoring?
What success looks like—and what failure would teach us
Success is reproducible prediction, reliable flow rates, manageable permitting, and competitive LCOE (levelized cost of electricity) compared to alternatives. Failure would mean models are overfit, subsurface complexity remains too high, or costs stay elevated. Both outcomes teach us. The industry must prefer rapid feedback loops: test fast, publish results, and adjust models. Will developers publish negative results or only successes?
Practical next steps for stakeholders
For investors: fund staged exploration portfolios and insist on milestone-based funding. For utilities: sign conditional offtake term sheets for demonstration projects. For policy makers: fund regional mapping and underwriting for early wells. For community groups: demand baseline monitoring and enforceable mitigation. Ask yourself: what would it take for you to say yes to a pilot project?
Closing perspective: pragmatic optimism
Zanskar’s announcement is a signal, not a conclusion. It mirrors long-running scientific work and points toward a practical pathway for unlocking heat that’s already underfoot. We should be neither euphoric nor dismissive. The right posture is pragmatic optimism: believe in the physics, test aggressively, protect communities, and align incentives so success is rewarded fairly. What will you commit to doing now that the risk profile has changed?
#Geothermal #HiddenHeat #AIforEnergy #EnergyInvestment #CleanBaseload #Geoscience #RenewableInfrastructure
Featured Image courtesy of Unsplash and Mike Bravo (lUyri-qNblA)