What if the next nuclear plant didn’t rise above the horizon at all? A US start-up called Deep Fission is betting that the safest—and potentially cheapest—place for a reactor could be more than a mile underground, inside the kind of vertical shaft normally drilled for oil, gas or geothermal projects.
The company’s concept, a modular reactor design known as Gravity, would sit about 1,600 metres below the surface, submerged in water inside a slim borehole. The idea is not to invent a brand-new kind of nuclear physics, but to change where the reactor operates—using the natural environment to do some of the work usually handled by thick steel, concrete and large containment structures.
How a nuclear waste insight became an “underground plant” concept
Deep Fission’s approach traces back to research into deep geological storage for nuclear waste. Scientists and engineers studying how to isolate spent fuel long-term have long focused on stable rock formations, high pressures and limited access—conditions that, in theory, can also support a compact reactor system.
Instead of only placing waste underground, Deep Fission proposes putting the entire reactor module down there. Gravity would operate as a pressurised water reactor using low-enriched uranium, similar to fuel and reactor principles used in many existing nuclear plants.
Why 1,600 metres matters: pressure and a mineral barrier
The engineering case hinges on what happens at depth. At around 1,600 metres, the water above the reactor creates a natural pressure of roughly 160 atmospheres (about 16 megapascals). Conventional pressurised-water reactors maintain comparable pressures with large, complex systems and heavy-duty pressure vessels. Gravity’s concept relies heavily on the water column and surrounding geology to provide that pressure.
Deep Fission also points to the surrounding rock as a kind of permanent, mineral “envelope.” Instead of a prominent above-ground containment building, the reactor would be encased within deep geological layers.
Claves
- Deep Fission proposes placing a modular reactor called Gravity about 1,600 metres underground.
- The design is based on a pressurised water reactor concept using low-enriched uranium.
- At depth, the water column provides roughly 160 atmospheres of natural pressure.
- The company says rock layers could serve as a geological containment barrier instead of massive surface domes.
- Challenges flagged include long-term monitoring, corrosion/material ageing, and complex maintenance logistics.
Using oil and geothermal know-how to speed up construction
A major pitch is that deep drilling is already a mature industrial capability. Oil and gas operators routinely drill kilometres down, and geothermal projects circulate water through hot rock formations. Deep Fission argues that this existing skill base could shorten build schedules compared with traditional nuclear construction.
According to the company’s timeline outline, a project could involve:
- About 4 weeks to drill the vertical shaft,
- Roughly 10 weeks to lower and install the reactor module and related equipment,
- Nearly 2 months of testing before connecting to the grid.
In total, Deep Fission targets operation in under six months on a prepared site—far faster than the multi-year timelines typical of large nuclear plants.
Safety case: geology as a “passive” layer of protection
Nuclear energy’s toughest political hurdle remains the fear of high-impact accidents. Deep Fission’s answer is to reduce what’s exposed at the surface. With the reactor positioned deep below ground, the company argues that severe scenarios would unfold far from population centres and outside the reach of many surface threats, including extreme weather.
The plan connects the underground reactor to surface power systems through a sealed column where hot water and steam rise to drive turbines. Supporters of the concept frame gravity and geology as additional safety factors that don’t rely on external power supplies or active systems.
However, the underground setting introduces its own unresolved questions, including long-term sensor reliability, communications robustness under high pressure and humidity, and how emergency and inspection procedures translate to a deep-shaft environment.
Costs, market targets and early interest
Deep Fission positions Gravity as a potentially lower-cost approach to nuclear power. The company has outlined an electricity cost target of €50 to €70 per megawatt-hour. It also says it has a pipeline of interested customers representing more than 12 gigawatts of potential demand, particularly in US states such as Texas, Utah and Kansas, where industrial energy demand and land availability are high.
On project economics, the company claims the approach could be up to 80% cheaper than building a large conventional plant—though it would still need to prove that deep installation and long-term operations do not introduce offsetting costs.
Critics point to maintenance, ageing materials and trust issues
Scepticism centres on practical realities. Major maintenance or refuelling would require complex lifting operations through a deep, narrow shaft—very different from servicing equipment in above-ground reactor buildings. Engineers also highlight the unknowns around corrosion and material ageing under sustained pressure and local groundwater chemistry conditions.
Beyond engineering, there is the issue of public confidence. A buried plant may reduce the visual footprint, but it could also raise concerns about oversight and transparency—two factors that strongly influence nuclear acceptance.
Where underground nuclear could fit in the energy mix
Gravity arrives amid a wider push to reinvent nuclear power through smaller, modular designs. In that crowded field—molten salt concepts, high-temperature gas reactors and waste-burning proposals—Deep Fission’s distinguishing move is largely about location and containment strategy, not fuel novelty.
If reactors like Gravity perform as proposed, they could be marketed as steady, compact power sources for grids increasingly dominated by intermittency from wind and solar, with potential applications ranging from heavy industry to data centres and hydrogen production.
