The timing is significant. Exploration is moving toward ultra-deep formations more than 8,000 meters beneath the surface and into increasingly complex deepwater environments, while unconventional resources account for much of the growth in reserves and production. At such depths, familiar assumptions begin to fail. Pressures can exceed 140 megapascals and temperatures can rise above 200 degrees Celsius, creating conditions in which rocks deform unpredictably, fluids behave in unfamiliar ways, and conventional seismic imaging loses accuracy. Meanwhile, hydrocarbons must be studied alongside carbon capture, utilization and storage, hydrogen, geothermal energy, artificial intelligence, and the emerging economic value of subsurface data.

The challenge list was launched in 2021 by the journal’s editorial board, which invited contributions from researchers and drew on strategic reports from organizations including the International Energy Agency, the American Association of Petroleum Geologists, and the Society of Petroleum Engineers. Candidate questions were assessed through expert workshops, external review, and screening criteria that included originality, urgency, transformative potential, interdisciplinarity, long-term influence, and balance across the field. Researchers from China University of Petroleum ultimately assembled a portfolio of 100 fundamental questions spanning exploration and development, storage and pipeline networks, refining and petrochemicals, petroleum materials, carbon management and emerging energy systems, and energy economics and digital transformation.

The authors classify the problems into three broad scientific patterns, revealing that petroleum research is no longer focused solely on finding and extracting oil and gas. Sixty-eight challenges are mechanism-oriented, seeking theories that explain why complex processes occur across multiple scales. These include the coupled effects controlling hydrocarbon accumulation in ultra-deep formations and the behavior of gas, liquid, and solid interfaces during multiphase catalysis. Thirteen are technology-oriented, addressing engineering limits such as high-precision geophysical prediction and drilling-fluid rheology under extreme pressure and temperature. The remaining 19 are system-level problems, involving the optimization of entire energy chains, the resilience of global supply networks, and the ownership and pricing of industrial data.

Among the ten challenges highlighted as particularly disruptive is the search for a unified theory of hydrocarbon accumulation and preservation below 8,000 meters. Solving that problem could clarify how temperature, pressure, mineral reactions, fluid migration, and rock deformation interact in the deepest reservoirs, potentially extending effective exploration toward 10,000 meters. Another question concerns the survival limit of liquid hydrocarbons in ultra-deep formations. If researchers can determine how organic molecules remain stable—or transform—at temperatures approaching 250 degrees Celsius, they could revise long-standing assumptions in organic geochemistry about where liquid petroleum can exist.

Several of the highlighted problems connect petroleum engineering directly to climate technology. One concerns water-free fracturing using supercritical carbon dioxide, a fluid state reached above its critical temperature and pressure in which carbon dioxide combines gas-like mobility with liquid-like density. Researchers must understand how this unusual phase changes during injection, how it transports proppant into fractures, and how it interacts with rock and formation fluids. A successful approach could reduce freshwater use in hydraulic fracturing while creating a pathway for carbon dioxide storage. The authors suggest that, if technical and economic barriers are overcome, such systems could help shift carbon capture and storage from a costly obligation toward a revenue-generating industrial process.

Another proposed breakthrough is the development of reservoir nanotracers capable of revealing the distribution of remaining oil at pore scale. These engineered particles would need carefully designed surfaces so they can travel through complex porous networks, resist chemical degradation, and produce detectable signals without becoming trapped prematurely. In principle, their movement could provide a form of “underground CT” imaging, allowing engineers to map fluid pathways that conventional well measurements cannot resolve. The resulting information could improve recovery strategies by showing where oil remains, how it is connected, and which microscopic channels control its movement through the reservoir.

The list also treats the subsurface as a potential platform for renewable energy and large-scale storage. In enhanced geothermal systems, engineers inject fluid into hot rock to create or reactivate fractures through which heat can be recovered. A central challenge is understanding the dynamic interaction between artificial fractures created by stimulation and natural fracture networks already present in the rock. Better control of these interactions could improve heat extraction while reducing the risk of induced seismicity, a major social and regulatory concern. For underground hydrogen storage, researchers must determine how hydrogen reacts geochemically and biologically with depleted reservoirs. Microorganisms may consume hydrogen, while minerals and formation fluids can alter its composition; controlling these losses could raise the working-gas ratio from below 50 percent to above 80 percent.

Ultra-long gravity heat pipes represent another proposed route to extracting heat from hot dry rock. These devices use phase changes in a working fluid to transport thermal energy over long distances, but their performance depends on evaporation, condensation, flow resistance, material stability, and pressure management at extreme temperatures. Identifying the limits of heat transfer and selecting suitable working fluids could lower the cost of geothermal electricity, with the authors estimating a possible levelized cost of 5 to 8 U.S. cents per kilowatt-hour. Such a target would place geothermal power closer to the cost range of other competitive low-carbon energy sources, although substantial engineering validation would still be required.

Artificial intelligence appears throughout the roadmap not simply as a tool for faster computation, but as a force changing how petroleum science is practiced. More than 30 of the challenges depend heavily on data, algorithms, or intelligent systems. Physics-informed neural networks can combine governing equations with observations, while differentiable programming allows models to be optimized through machine-learning methods. Neural operators can learn relationships between fields, such as pressure, temperature, and saturation, across changing geological conditions. Together, these techniques could produce “grey-box” models that retain the flexibility of data-driven systems while preserving some physical interpretability. Digital twins would extend this approach by continuously updating virtual representations of reservoirs as new measurements arrive, potentially turning reservoir management from periodic, experience-based intervention into real-time adaptive control.

The roadmap also identifies a rapidly expanding frontier in energy economics and global infrastructure. Satellite observations, vessel tracking, and machine-learning forecasts could enable daily monitoring of tanker capacity and freight rates, offering an independent view of oil and gas transportation markets. At the same time, subsurface data—once treated mainly as technical records—could become a factor of production with measurable ownership rights, economic value, and pricing mechanisms. The authors argue that future research must establish how geological information is created, shared, protected, and monetized. More broadly, the 100 questions reflect a transition from petroleum as a discipline of resource extraction to what the authors call “smart integrated energy science,” combining subsurface storage, oil, gas, hydrogen, electricity, heat, carbon-cycle management, and intelligent decision-making. The list is deliberately open-ended: as researchers solve existing problems, new uncertainties will emerge. Its central message is that defining what remains unknown may be the first step toward transforming the energy system.

Subject of Research: Not applicable

Article Title: One hundred grand challenges in petroleum science.

Web References: https://doi.org/10.1016/j.petsci.2026.06.013

Image Credits: Xiao, L. Z., Jin, Y., Zhang, L. B., et al. (journal cover)

Keywords: petroleum science, ultra-deep reservoirs, artificial intelligence, digital twins, carbon capture and storage, geothermal energy, underground hydrogen storage, supercritical carbon dioxide fracturing, reservoir nanotracers, energy transition

Hydraulic fracturing, widely known as ‘fracking,’ involves injecting fluid at high pressure into subterranean rock formations to create fractures, thereby enhancing the permeability of reservoirs and facilitating the flow of oil and gas. However, in naturally fractured reservoirs—where pre-existing fractures and faults disrupt the rock fabric—predicting the behavior of induced fractures becomes profoundly challenging. The interaction between induced hydraulic fractures and existing natural fractures plays a pivotal role in determining the efficacy of stimulation operations. The study by Gao, Zhou, Han, and colleagues methodically investigates this interaction, using advanced modeling techniques combined with empirical data.

One of the central challenges the research addresses is how natural fractures influence the geometry and propagation pathways of hydraulic fractures. Traditional hydraulic fracturing models often assume a homogeneous rock medium, neglecting the heterogeneous fracture networks typical in many reservoirs. Such simplifications risk inaccurate predictions of fracture propagation, leading to inefficient reservoir stimulation and unexpected environmental consequences. By integrating the impact of natural fractures into their models, the authors have provided a more realistic framework for fracture propagation analysis.

The research utilizes a sophisticated numerical model capable of simulating the initiation and growth of hydraulic fractures in the presence of natural fracture networks. The model captures parameters such as fracture toughness, fracture aperture, and fluid pressure, alongside the mechanical properties of the rock matrix. This approach enables the observation of various fracture propagation scenarios, including intersection with natural fractures, deflection, termination, or crossing. Such comprehensive simulations reveal the stochastic and anisotropic nature of fracture growth in fractured reservoirs.

One of the most striking findings from the study is that natural fractures can serve both as barriers and conduits for hydraulic fractures, depending on their orientation, mechanical properties, and stress regimes. When hydraulic fractures encounter natural fractures aligned favorably with the local stress field, they may coalesce, resulting in extended fracture networks that significantly improve reservoir permeability. Conversely, unfavorable alignments may cause hydraulic fractures to arrest or divert, potentially reducing the stimulated reservoir volume.

Furthermore, the study emphasizes the critical role of in-situ stress anisotropy on the fracture propagation patterns. Stress fields within the Earth are rarely uniform, and variations in maximum and minimum principal stresses govern fracture initiation and growth directions. The interplay between these stress anisotropies and the spatial distribution of natural fractures results in complex fracture geometries that must be accounted for to optimize hydraulic fracturing designs. This insight underscores the necessity for detailed site characterization prior to fracturing operations.

Additionally, fluid injection parameters such as rate, viscosity, and volume play a substantive role in fracture propagation. The research demonstrates that high injection rates and low-viscosity fluids tend to promote fracture branching and penetration through multiple natural fractures, thus creating an intricate fracture network. In contrast, slower injection rates with higher-viscosity fluids produce more planar fracture geometries. These findings have direct implications for field operations since fluid properties can be engineered to tailor fracture geometries according to reservoir and production goals.

The coupling of mechanical behavior of rock and fluid flow through fractures is another sophisticated aspect of the model. Natural fractures often act as fluid conduits, so their interaction with newly formed hydraulic fractures affects not only structural integrity but also fluid migration patterns. Understanding these dynamics can lead to improved predictions of proppant placement, fracture conductivity, and long-term reservoir performance. The study’s integrated approach combining geomechanics and fluid dynamics is a significant advancement in fracturing science.

Importantly, the findings of this work have implications for environmental safety and risk management. Uncontrolled propagation of hydraulic fractures into unwanted zones or aquifers is a major concern. By understanding how natural fractures dictate fracture pathways, operators can better anticipate and mitigate risks related to induced seismicity, groundwater contamination, and surface impacts. The authors advocate for adaptive fracturing designs informed by detailed fracture network characterization and in-situ stress measurements as a pathway to safer operations.

The research also sheds light on the temporal aspects of fracture growth. Hydraulic fractures rarely propagate instantaneously but evolve dynamically with changing fluid pressure and rock stresses. The study’s time-dependent simulations provide insights into fracture propagation speed, interaction duration with natural fractures, and post-injection fracture closure behavior. These temporal factors influence proppant placement efficacy and reservoir productivity, highlighting the importance of temporally resolved modeling in fracturing operations.

Moreover, the authors discuss the scale dependency of fracture propagation mechanisms. While laboratory experiments provide critical data on fracture initiation, reservoir-scale phenomena involve complex fracture interactions across multiple length scales, from micro-fractures to kilometers-long faults. The multi-scale modeling methodology employed bridges laboratory results with field-scale applications, offering a practical tool for engineers and geoscientists to predict and control hydraulic fracture geometry.

Another intriguing element is the influence of natural fracture aperture and roughness on induced fracture propagation. Narrow or rough natural fractures may act as partial barriers, allowing some fluid penetration but reducing the mechanical coupling between fractures. The study indicates that aperture variability leads to heterogeneous pressure distribution within fractures during fluid injection, which in turn affects propagation direction and fracture branching. These microstructural details are essential to capture for high-fidelity reservoir stimulation models.

Furthermore, fracture closure and proppant embedment after fluid injection cease are vital for maintaining fracture conductivity over reservoir life. The authors provide insights into how interaction with natural fractures both aids and complicates fracture propping. In some cases, natural fractures may provide additional conductivity pathways that remain open even after hydraulic fractures begin to close. This phenomenon could enhance long-term production but requires careful management of fracture design and proppant selection.

To sum up, the study by Gao and colleagues establishes a comprehensive framework for understanding hydraulic fracture propagation in the naturally fractured reservoir setting, which has long eluded precise characterization. This research represents a pivotal advance in the hydraulic fracturing field, combining rigorous numerical simulation with geological complexity to improve stimulation strategies. As resource extraction becomes increasingly challenging, such scientific breakthroughs will be indispensable for achieving sustainable and efficient energy development.

As we look to the future, integrating real-time monitoring systems with the predictive capabilities presented in this research could open new frontiers in hydraulic fracturing. Coupling microseismic monitoring, fiber optic sensing, and advanced geophysical imaging with fracture growth models could provide operators immediate feedback to optimize fracturing parameters adaptively. This fusion of technology and geology will undoubtedly drive more intelligent reservoir exploitation, reducing costs and environmental footprints.

In conclusion, the intricate dance between hydraulic and natural fractures governs the success of stimulation treatments in fractured reservoirs. Gao et al.’s thorough investigation elucidates many previously underappreciated factors, offering a robust toolset for industry practitioners and researchers alike. As the global energy landscape continues to evolve, such integrated studies reinforce the importance of multidisciplinary research in tackling the complex problems at the heart of unconventional resource development.

Subject of Research: Hydraulic fracture propagation influenced by natural fractures in fractured reservoirs.

Article Title: Study on hydraulic fracture propagation under the influence of natural fractures in fractured reservoirs.

Article References:
Gao, Q., Zhou, Z., Han, Y. et al. Study on hydraulic fracture propagation under the influence of natural fractures in fractured reservoirs. Environ Earth Sci 84, 425 (2025). https://doi.org/10.1007/s12665-025-12428-5

Image Credits: AI Generated