Not long ago, “enhanced geothermal systems” (EGS) — sometimes just called geothermal fracking — felt like a science-fair cousin of shale. Lately, though, they keep showing up in hallway chatter at SPE events, and, more tellingly, inside board-package slide decks. The U.S. Department of Energy’s (DOE's) GeoVision model even suggests the technique could unlock roughly 60 GW of steady, carbon-free capacity by mid-century. That projection is hardly carved in granite, yet it’s large enough to make reservoir engineers sit up and rerun the math.

Why the sudden buzz? For starters, the basic playbook looks familiar to anyone who has steered a rotary steerable system through tight carbonate windows: drill into hot but stubbornly impermeable rock, create a fracture network with high-pressure water, then circulate fluid to haul the heat topside. In other words, the same skill set that cracked open the Barnett and Permian may crack open crustal heat as well. Service companies already own the horsepower — pumps, proppant logistics, fiber-optic diagnostics — while field crews know how to stage-frac, monitor microseismic and handle live-well pressures.

Still, opportunity is not destiny. Unlike shale wells that pay back in three frenzied years, EGS projects have to keep fractures open and conductive for decades, all while avoiding runaway induced seismicity or silent water loss into neighboring faults. Mineral scaling, too, could gum up production tubing faster than expected if fluid chemistry drifts. Early data from the DOE’s FORGE site in Utah appears to show that careful stress-shadow management and closed-loop circulation mitigate some of these risks, yet commercial proof remains thin.

From a market lens, firm geothermal power dovetails with corporate Scope 1 targets and with utilities desperate for around-the-clock generation that isn’t coal, nuclear or a vast battery farm. Whether that alignment translates into bankable offtake agreements at scale is, frankly, still up for debate. But the mere fact that oil-field veterans and power-selector analysts now share conference panels hints that a crossover moment may be underway — one where unconventional completions know-how shoulders its way into the low-carbon conversation.

Lessons from the shale revolution

When petroleum engineers first poke around the enhanced-geothermal literature, they often do a double-take: the maps show heat nearly everywhere, yet the well tests are anything but prolific. The catch is circulation, not temperature. Granite may sit at 200° C 3 km down, but it’s tough to coax that heat to the surface unless one carves out a high-surface-area flow path that stays open for decades.

That challenge doesn’t scale the shale crowd, though. Horizontal motors that keep a bit in the pay to within a yard — or, more accurately, within a meter — now hold up surprisingly well at 200° C after tool suppliers swapped out elastomers and beefed up downhole electronics. The preferred layout looks something like a racetrack: a vertical or slightly deviated injector meets a production leg 400-plus meters away, down-dip enough to avoid thermal short-circuiting but close enough that engineers are not burning horsepower on every gallon.

The frac itself feels similar, yet not identical. Crews still pump slickwater at 30 bbl/minute to 40 bbl/minute, but proppant loads get dialed back because crystalline basement tends to prop itself on rough fracture faces. Sand, if overdone, also invites silica scaling, which is the bane of long-term injectivity. Stage timing gets more surgical, too. Operators hesitate to stack net pressure across a fault block; they’ll often finish two stages, throttle back and let the microseismic array tell them if they’re flirting with a felt event. On-the-fly adjustments, guided by fiber DAS and buried geophones, are now routine rather than experimental.

Is the recipe working? The Utah FORGE test pad offers an encouraging, if not yet bankable, data point. In 2024 the team linked eight slick-water stages in injector 16A(78)-32 to four stages in a nearby producer at 2.5 km and 221° C. Injectivity improved by several-fold; tracer sweeps hinted at residence times consistent with roughly 5 MW of firm power per well pair, numbers that start to look like a proper field, not a pilot. Skeptics note that scaling and stress cycling over 20 years remain unproven, yet the test does not suggest the conventional tool kit translates far better to hot-dry-rock than earlier air-drilled experiments implied. The next couple of commercial debuts, likely sited on the Nevada Test Range and in West Texas, should tell us whether FORGE was a lucky one-off or the opening of a new play.

Problems of permeability and seismicity

Ask any completion engineer who has tried to keep a granite heat-exchanger alive past the first winter: the rock always answers back. Cool water shrinks the fracture walls, pore pressure rises, and before long the once-roomy flow paths feel tight enough to squeeze a credit card. Core tests say permeability can drop by 50% after a few hundred thermal swings, yet those same tests were cut from tidy cylinders in a lab, not a busted-up rhyolite laced with drilling mud. Field reality is messier — and, strangely, more forgiving.

Most operators now start by running a coupled thermo-poro-elastic model that looks eerily like a shale simulator with a heat-transfer sidecar. The math helps, but it still leaves big question marks. Should one spend extra pump horsepower to inject at 60° C instead of 40° C and slow the drawdown, or accept faster cooling and bet on cheap refracs later? The answer depends on local tariffs, turbine efficiency curves, even how much downtime the grid can tolerate if flow is rotated among doublets. At Newberry Volcano in Oregon, the team shut-in an overstressed producer for six months and saw injectivity bounce back to 70% of its start-of-life value — a pleasant surprise that may not repeat in drier, lower-permeability granites a few hundred miles south. Results like these suggest “rest cycles” could become as normal as workovers, but the jury is still out.

Seismicity brings a different kind of headache, the sort that lands project managers in county-commission meetings. Site traffic-light protocols in Utah and Nevada typically invoke mandatory shut-downs at moment magnitude about 3.0, with advisory actions beginning near the mid-2 range. Fiber DAS strings now flag sub-M₀ events in real time, but prediction remains an exercise in probabilistic humility. Pre-conditioning faults with low-rate acid or carbon dioxide seems to smooth pressure spikes — some researchers are evaluating gentler ramp-up schedules and pressure-managed step-rate injections to limit moment release, but field results remain preliminary.

So we muddle forward: forecast, pump, listen, tweak. It is an iterative, almost conversational workflow — one where geomechanics, heat transfer and community relations share the same call sheet. For now, the discipline’s best guess is that a doublet flowing 50 km to 80 km per second at 180° C can stay on line for two decades without rattling the neighbors, provided every stage is planned with a genuine fear of the unknown. That cautious optimism, rather than any single breakthrough, is what’s gradually making EGS bankable.

Following the money: Pilots are morphing into power plants

Until recently, the path from “cool demo well” to megawatt-scale revenue felt murky. That fog finally began to lift in 2025 — and, interestingly, the breakthrough had little to do with better drill bits or slicker reservoir models. Instead, geothermal plants placed in service after 2024 and before 2033 (or until the grid hits the law’s emissions trigger) qualify for the full section 45Y production tax credit. By most banker models, that single line of policy trims the levelized cost of energy by roughly 20%. Not a silver bullet, perhaps, but enough to pull spreadsheets from the red into the pale green.

Wall Street still wanted a long-term buyer, though. The surprise came from the data-center crowd, which burns through electricity like a jetliner drinks Jet-A. In June, Meta signed a long-term agreement to support the development of 150 MW of advanced geothermal capacity; commercial teams were not disclosed, but the current levelized cost is pegged to similar projects at roughly $80/MWh — remarkably close to what solar-plus-battery combos fetch in the same transmission zone. The contract’s length matches the slow thermal decline expected in an EGS reservoir, giving lenders the comfort they need to issue debt at pipeline-style leverage ratios.

Still, a few caveats loom. Section 45Y’s value is indexed to inflation; if Congress tweaks the formula, margins could erode overnight. And while Meta’s deal is encouraging, it hardly guarantees a queue of eager offtakers. Some utilities remain skeptical, quietly noting that EGS wells have not yet proven 30-year mechanical integrity at scale. They have a point: tracer tests tell a promising story, but rock mechanics can deliver unpleasant plot twists after a decade of heat drawdown.

From the service-company vantage, however, momentum feels tangible. Rigs and frac spreads idled by basin slowdowns can migrate to geothermal pads with minimal retooling, and crews appreciate the year-round work that isn’t hostage to $70-WTI mood swings. Put together — policy carrots, credit-worthy buyers and an under-utilized shale supply chain — EGS may finally have the ingredients for a commercial recipe. Whether the dish scales beyond a few showcase meals will hinge on how the first tranche of 2025-vintage projects ages. Drillers, investors and regulators alike would be wise to keep their risk models sharp and their expectations humble. The subsurface, after all, does not read incentive tables.