Ultra-Deep Geothermal Drilling & The Rise Of Black Swan Risks

Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!

Last Updated on: 19th March 2025, 06:15 pm

Deep drilling isn’t optional for enhanced geothermal systems (EGS), it’s the whole point. To understand why, think of the Earth’s crust as a hot soup. Near the surface, it’s merely lukewarm, barely useful beyond warming your house if you’re lucky. Go deeper, however, and temperatures rise rapidly, roughly 25 to 30 degrees Celsius for every kilometer drilled, though that varies wildly depending on geology. Accessing sufficient heat, ideally around 200–400 degrees Celsius to economically generate electricity, usually means reaching depths between 4 and 10 kilometers, often in tough, unforgiving rock.

As a note, this is one in a series of articles on geothermal. The scope of the series is outlined in the introductory piece. If your interest area or concern isn’t reflected in the introductory piece, please leave a comment.

Historically, we’ve only scratched the surface. The Kola Superdeep Borehole in Russia, drilled over two painstaking decades between 1970 and 1990, reached about 12 kilometers deep. It was an extraordinary feat, but one marred by harsh realities: drill bits frequently failed, and the deeper they went, the hotter and more plastic the rock became. At around 180 degrees Celsius, the borehole started deforming like a squeezed plastic straw, marking the limits of conventional drilling.

The oil and gas industry has surpassed this depth, at least on paper, drilling horizontally and vertically to depths exceeding 12 kilometers. However, these wells, like those in Qatar or Russia’s Sakhalin region, navigate softer sedimentary formations and avoid the scorching temperatures of deep geothermal targets. Iceland’s Deep Drilling Project (IDDP), by contrast, plunged straight into supercritical conditions at around 450 degrees Celsius just 4.5 kilometers down, proving both potential and peril. Their casings corroded swiftly, underscoring the limits of existing technology.

Enter novel drilling approaches promising to rewrite these rules — each fascinating, expensive, and accompanied by a healthy dose of skepticism. Take millimeter-wave drilling, championed by Quaise Energy, spun out from MIT, sitting at Technology Readiness Level (TRL 4 with 9 being commercialized). Instead of grinding rock, Quaise melts it using microwaves beamed downhole through specialized waveguides. Quaise claims this can reach depths of 20 kilometers with costs scaling linearly — not exponentially.

The catches? Their biggest lab test saw a 2.5 cm hole 2.5 m long, which is about a 4,000th of their claims for how deep they can go. As an engineering rule of thumb, you have to get to quarter-scale prototypes to be in the same physics ballgame, so they have a lot of scaling to do. Their vision of achieving a cost of roughly a thousand dollars per meter sounds optimistic at best and fantastical at worst. Real-world rock has fluids, fractures, and surprises, and microwaves notoriously struggle in wet environments. And last but not least, what happens to the melted rock? They ran compressed air to the bottom of the 2.5 m hole and it blew the rock out as thin threads, but getting air to blow melted rock several kilometers straight up strikes me (and an awful lot of other people) as deeply unlikely. It’s much more likely to stick to equipment and the sides of the hole and gum up the works. Still, if Quaise can keep the microwaves from scattering and overheating components deep underground, it could transform EGS economics.

GA Drilling’s PLASMABIT (TRL 4-5) follows a parallel path, using plasma torches to thermally fracture rock. Their lab tests show rock fractures beautifully under extreme heat, but downhole conditions — pressurized water, corrosive environments, unpredictable rock compositions — are harsher. GA hedged their bets with incremental advances like their AnchorBit, essentially a downhole stabilizer, already demonstrating success at boosting conventional drilling rates in lab settings. But scaling plasma fracturing tools to field-ready depths remains technically daunting. Imagine igniting and maintaining a plasma torch kilometers beneath your feet — any malfunction could turn expensive quickly. People I know who have worked with plasma torches, including chemical processing Paul Martin, make it clear that they are hard to control.

Other techniques, such as thermal spallation, employ intense heat jets to flake away rock, promising drilling speeds significantly faster than conventional methods. Potter Drilling (TRL 5) and the EU ThermoDrill project (TRL 6-7) demonstrated promising penetration rates in lab and small field trials. Yet, there’s a critical caveat — this approach hinges on rock types cracking predictably under thermal stress. Encountering non-cooperative geology, like softer rocks that melt rather than spall, could send costs skyrocketing. And when rock is much hotter and more plastic as it is down deep, this is unlikely to perform nearly as well.

High-power laser drilling also flirts with transformational claims. Labs have shown lasers easily slicing through shale and sandstone, but delivering a coherent, intense beam several kilometers underground isn’t trivial. Lasers need perfectly engineered optics and fiber cables resistant to immense pressure and heat. Real-world demonstrations have been limited, and any water in the rock can scatter the laser beam, dramatically reducing efficiency. Laser-assisted drilling is intriguing, perhaps even viable in certain conditions, but far from proven at depth.

Traditional mechanical drilling isn’t idle. Hammer drilling technologies, now at TRLs around 6 or 7, are beginning to reliably demonstrate higher penetration rates and greater durability in hard crystalline rock at moderate depths. Polycrystalline diamond compact (PDC) bits, reaching TRLs of 8 or higher, have significantly increased drilling efficiency in tough geological conditions, reducing downtime due to frequent bit replacements. Directional drilling, well-established at TRL 9, allows precise targeting of geothermal reservoirs, optimizing resource access and minimizing drilling lengths. The primary strength of these approaches lies in their proven operational history and incremental improvements that reduce risk relative to radically new methods.

However, mechanical drilling remains challenged at depths beyond 7 kilometers due to increasing temperatures that degrade tool integrity and rock becoming less brittle and more plastic, making efficient drilling increasingly difficult. The key technical risks include managing extreme heat, minimizing bit wear, and avoiding catastrophic tool failures that can quickly escalate project costs. Even incremental improvements here might yield better returns than betting everything on entirely novel methods.

This brings us back to why ultra-deep drilling is tricky. Below certain temperatures, rock becomes ductile — less prone to fracturing and more likely to deform and seal any induced fractures. Fracking can temporarily induce fractures, but maintaining long-term permeability remains unproven. Moreover, ultra-deep drilling means operating at the extremes of material capabilities: casing steels weaken, electronics fail, and unexpected geologic surprises, such as overpressured fluids or even magma, can turn a promising project into a costly dead-end overnight.

Given this, deep geothermal drilling epitomizes what’s known as a ‘long-tail risk,’ or as Bent Flyvbjerg vividly frames it — a classic breeding ground for ‘black swan’ events. These unpredictable, rare, and high-impact outcomes aren’t merely theoretical—they stack up alarmingly when combining extreme depths, first-of-a-kind (FOAK) technologies, and unprecedented geological conditions. Each added kilometer doesn’t just increase capital costs; it exponentially multiplies uncertainties, creating layers of technical, geological, and economic risks. Novel drilling methods magnify this uncertainty: technologies that function beautifully in controlled laboratory settings can falter disastrously under harsh, real-world conditions deep underground. Flyvbjerg’s insights warn us that optimism bias frequently underestimates the complexity and potential for catastrophic failure in such innovative ventures, making deep geothermal drilling a compelling but perilously uncertain endeavor.

Fly too close to the Earth’s molten heat, and your investment can evaporate — quite literally — if you hit supercritical conditions unprepared. Thus, novel drilling technologies, while alluring, must navigate a perilous path: proving they can truly lower costs, reliably manage surprises, and achieve consistent economic performance at industrial scale.

The uncomfortable truth is that deep geothermal drilling — particularly using cutting-edge, largely untested methods — embodies exactly the type of long-tail, black-swan-rich endeavor that Bent Flyvbjerg has shown is most susceptible to massive delays, cost overruns, and outright failures. Betting heavily on these ambitious but immature technologies might yield revolutionary breakthroughs, or just as likely, become another cautionary tale of expensive hubris chasing dreams far below ground. My opinion that geothermal for electrical generation would remain a rounding error globally hasn’t changed after going deep on advanced drilling technologies.