Fractured Rocks Hold the Key to Unlocking the Energy Transition Underground

The invisible infrastructure of the energy transition may not be wind turbines or solar farms but the cracks, joints and faults threading through rocks hundreds of metres below our feet. That is the central argument of a Perspective article published on 9 December 2025 in the journal GeoEnergy Communications, in which Qinghua Lei and Chin-Fu Tsang contend that understanding coupled processes in fractured media is a make-or-break frontier for decarbonisation. The authors frame fractures as simultaneously the enablers and the bottlenecks of five subsurface technologies: geothermal energy, carbon sequestration, hydrogen storage, nuclear waste disposal and critical mineral extraction.

The unifying concept is the coupling of thermal, hydrological, mechanical, chemical and biological effects, abbreviated THMCB. Thermal stresses open or close fracture apertures; pore pressures from fluid injection can trigger slip on faults; chemical reactions dissolve minerals and enlarge flow paths or precipitate new phases that clog them; and microbial activity alters mineral surfaces and promotes biofilm growth that modifies fracture geometry. These feedbacks are not side effects, the authors argue, but the governing dynamics of fractured rock, and they evolve dynamically across scales, creating major challenges for prediction and management.

Geothermal energy illustrates the double-edged nature of fractures most vividly. In Enhanced Geothermal Systems, engineers inject water into low-permeability hot rock to create or extend fracture networks, effectively turning the rock mass into a giant heat exchanger between injection and production wells. Without sufficient fracture connectivity, heated water flows too slowly for viable energy production. Yet cold-water injection induces thermal contraction and elevated pore pressures that can propagate new fractures or reactivate dormant ones, and fluid injection inevitably carries the risk of induced seismicity. In several documented cases, large induced earthquakes have led to the suspension of EGS projects. Fractures can also short-circuit flow between wells, letting water reach the production well before it has extracted enough heat.

Carbon sequestration faces a parallel dilemma. The strategy involves injecting supercritical CO2 into deep saline aquifers or depleted oil and gas reservoirs and trapping it there for centuries or longer, with a secure, low-permeability caprock essential to prevent buoyant CO2 from migrating upward. Fractures in the target reservoir can enhance injectivity and storage capacity, but injection-induced pressure increases may then transmit rapidly over long distances. Fractures cutting the caprock could act as leakage conduits, so operators must limit injection rates and pressures to avoid hydro-fracturing the caprock or reactivating faults. Injected CO2 acidifies brine and reacts with minerals over time, slowly altering fracture apertures, while salt precipitation can locally clog fractures. Mineralisation, which locks carbon in solid form, proceeds rapidly in reactive basalts but much more slowly in typical saline aquifers.

Hydrogen storage, envisioned as a way to convert excess renewable electricity into a storable fuel, confronts similar risks amplified by cycling. Seasonal to daily injection and withdrawal of hydrogen subjects reservoirs and caprocks to repeated pressure and stress swings that could cause rock fatigue, with attendant risks of induced seismicity and land subsidence or uplift. Hydrogen may also stimulate microbial or geochemical reactions that generate corrosive byproducts or consume the stored gas, with fractures offering pathways for such processes. Salt caverns, already proven for natural gas storage, rely on the tightness and self-healing behaviour of salt, making the avoidance of fracture creation in cavern walls during pressure cycling the central challenge. Lined rock caverns, considered promising for small to intermediate-scale storage, face fracture-induced concrete cracking and hydrogen embrittlement of steel linings.

Nuclear waste disposal adds a millennial dimension. The accepted concept seals canisters in tunnels excavated hundreds of metres underground in crystalline formations, salt domes, bedded salt or claystones, with sites chosen to minimise connected fractures that could carry groundwater through the repository. Fractures remain unavoidable, however: excavation redistributes stress and creates a damaged zone of induced fractures around tunnels, raising permeability near openings, while heat from spent fuel can generate thermal cracking in brittle rock or drive drying and swelling cycles in clays that open gaps at bedding planes. Over such timescales, future glacial cycles and earthquakes may impose significant stresses, so performance assessments must account for coupled THMC-mediated radionuclide transport.

Even mining fits the framework. Many critical minerals, including lithium, cobalt, nickel, copper and rare earth elements, come from deposits formed by hydrothermal fluids circulating through fracture networks; porphyry copper ore veins are frequently former fractures infilled with minerals. Numerical models of coupled deformation and fluid flow now guide exploration in structurally complex terrains, while extraction itself demands managing fracture-controlled water inflows, circulating leaching solutions through fracture networks, and guarding against fault slip, tunnel collapse and acid mine drainage.

The authors identify five key challenges. Upscaling laboratory measurements to field scale is difficult because collective behaviours at large scale may not manifest in small-scale studies. Uncertainty is pervasive: fracture density, aperture and connectivity vary widely, and poorly constrained fault frictional strength can push predicted induced seismicity from negligible to project-threatening, demanding rigorous probabilistic assessment rather than after-the-fact error bars. Induced seismicity remains the most publicly visible risk, with forecasting of timing and magnitude still elusive and traffic-light mitigation systems often reactive. Maintaining long-term integrity of wellbores, caprocks, caverns and repositories against slow coupled degradation is another concern, as is modelling full 3D coupled systems, which remain computationally demanding and often rely on simplifications that risk neglecting key processes.

The proposed path forward leans on emerging tools: digital twins that ingest downhole pressure, microseismic and surface deformation data to update coupled simulations in real time; physics-informed machine learning that embeds governing equations into learning processes; distributed fibre-optic sensing, microseismic arrays and geodetic monitoring; exascale computing and new pressurised true-triaxial laboratory cells that image fracture development in real time; and underground pilot projects. The authors also trace the field’s institutional history, from the DECOVALEX project launched in 1992 for nuclear waste disposal benchmarking to the CouFrac conference series begun in Wuhan in 2018, with the next meeting planned for Uppsala in 2026. Ultimately, they conclude, progress will require integration across disciplines and transparent engagement with society, transforming fractured media from a source of uncertainty into a foundation for safe and scalable subsurface energy systems.

Subject of Research: Coupled THMCB processes in fractured subsurface media and their role in geoenergy applications for the energy transition

Article Title: Coupled processes in fractured media: a key to the energy transition

Article References: Lei, Q., & Tsang, C.-F. (2025). Coupled processes in fractured media: a key to the energy transition. GeoEnergy Communications, 1(1), Article 11. https://doi.org/10.1007/s44421-025-00011-4

Image Credits: AI Generated

DOI: 10.1007/s44421-025-00011-4

Keywords: fractured media, coupled processes, THMCB, geothermal energy, carbon sequestration, hydrogen storage, nuclear waste disposal, critical minerals, induced seismicity, digital twins, uncertainty quantification, upscaling