One hundred grand challenges in petroleum science

The advancement of science often begins with the identification and articulation of key questions. In 1900, the German mathematician David Hilbert presented 23 mathematical problems at the International Congress of Mathematicians in Paris, charting the course for 20th-century mathematics and influencing algebra, geometry, analysis, and many other branches. More than a century later, petroleum science—a discipline that underpins modern industrial civilization—faces a similar "problem-driven" moment.

In its August 2026 issue, Petroleum Science officially releases the "100 Grand Challenges in Petroleum Science," a landmark editorial five years in the making that is poised to exert a lasting influence on the development of petroleum science for decades to come.

Petroleum science today stands at a historic turning point. Exploration is pushing into ultra-deep (>8,000 meters) and deepwater frontiers, unconventional resources have become the dominant source of reserve and production growth, and carbon neutrality and artificial intelligence (AI) are reshaping both the industrial landscape and the boundaries of the discipline. Classical accumulation theories and geophysical techniques fail in ultra-deep, ultra-high-pressure (>140 MPa), and ultra-high-temperature (>200 °C) settings; the mechanisms of multiphase flow in nanopores and the causes of rapid production decline in unconventional wells remain poorly understood. Furthermore, global climate governance is placing an imperative on petroleum science to integrate with emerging fields such as carbon capture, utilization and storage (CCUS), hydrogen energy, and geothermal energy.

In 2021, the editorial board of Petroleum Science launched an initiative to solicit the "100 Grand Challenges in Petroleum Science" from the global community. These include strategic reports from the International Energy Agency (IEA), the American Association of Petroleum Geologists (AAPG), the Society of Petroleum Engineers (SPE), among others. Through iterative expert workshops, broad external review, and multi-criteria screening—evaluating each candidate against scientific originality, practical urgency, transformative potential, interdisciplinary nature, temporal impact, and disciplinary balance—a team of researchers from China University of Petroleum  compiled a final list of 100 fundamental scientific questions.

These challenges span six domains: Exploration and Development (30 challenges); Storage, Transportation, and Pipeline Networks (15); Refining and Petrochemicals (15); Petroleum Materials Science (10); CCUS–Geothermal–Hydrogen (15); and Energy Economics and Digital Transformation (15).

In-depth analysis of the  challenges also revealed three archetypes. Mechanism-oriented problems (68 items) seek unified theories of multi-scale coupling, forming the "foundation" of the entire knowledge system by answering the fundamental question of why phenomena occur—for example, the multi-factor coupling theory of ultra-deep hydrocarbon accumulation or the unified theory of gas-liquid-solid interfacial transfer in multiphase catalysis. Technology-oriented problems (13 items) push engineering limits under extreme conditions, representing boundaries such as the limits of high-precision geophysical prediction for ultra-deep reservoirs or rheology control of drilling fluids under ultra-high temperature and pressure. System-level problems (19 items) address full-chain integration and energy transition, dealing with how to integrate and optimize under complex constraints—for example, the full-chain efficiency of CCUS, the evolution of global oil and gas supply chain resilience, and the value and pricing mechanism of oil and gas data.

From this list, the authors have distilled the ten most disruptive challenges:

• Multi-factor coupling theory of hydrocarbon accumulation and preservation in ultra-deep formations (>8,000 m)—a challenge that could extend effective exploration depth toward 10,000 m and dramatically expand global hydrocarbon resources; 
• Control mechanisms for the survival limit of liquid hydrocarbons in ultra-deep formations, potentially rewriting organic geochemistry by extending the liquid window to 250°C or deeper;
• Phase-transition control and proppant transport enhancement mechanisms of supercritical CO₂ water-free fracturing, which would enable zero-freshwater fracturing and transform CCUS from a cost center into a profit center;
• Surface and interface design of reservoir nano-tracers, which could achieve "underground CT" imaging of pore-scale remaining oil distribution;  
• Dynamic interaction between artificial and natural fracture networks in enhanced geothermal systems (EGS), which could remove the social-acceptance barrier of induced seismicity and enable safe commercial geothermal energy;
• Mechanisms of hydrogen geochemical and microbial consumption in depleted reservoir storage, a decisive step for a commercial hydrogen economy that could increase the working gas ratio of depleted-reservoir hydrogen storage from below 50% to above 80%;
• Working-fluid selection and phase-change heat transfer limits in ultra-long gravity heat pipes for hot dry rock heat extraction, which could reduce the levelized cost of geothermal electricity to USD 0.05–0.08 per kilowatt-hour;
• Real-time monitoring of global tanker fleet capacity and freight rate forecasting based on multi-source satellite data, establishing the first daily independent inventory monitoring system;
• Value composition, ownership, and pricing mechanism of oil and gas subsurface data, elevating data from "records" to a "factor of production"; and
• Real-time construction and evolution theory of subsurface reservoir digital twins based on data-mechanism fusion, which would transform reservoir management from offline empirical operation to real-time adaptive optimization.

A main insight of this work is that AI has evolved from an enabler into a core driver. More than 30% of the challenges rely on data, algorithms, or intelligent systems as key solution pathways. Specifically, AI is:

• Changing the cognitive style of petroleum science through the convergence of data-driven and mechanism-driven approaches; 
• Reshaping the research chain through digital twins that allow high-fidelity simulation in virtual space; and 
• Redefining value standards as data becomes an independent factor of production. Physics-informed neural networks (PINNs), differentiable programming, and neural operators represent important directions for future computational petroleum science, embedding physical knowledge into algorithmic structures to create "grey-box" models that retain the flexibility of data-driven methods while providing partial explainability.

Notably, low-carbon transition is reflected in more than 30 challenges covering CCUS, geothermal energy, hydrogen energy, biomass conversion, and low-carbon refining. This transition is reconstructing the discipline's value standards—oil and gas assets must now account for carbon risk, stranding risk, and compliance cost; its time scales—CCUS and underground hydrogen storage extend prediction horizons to centuries or millennia; its spatial scales—from global carbon cycles to molecular CO₂-mineral interfacial reactions; and its disciplinary boundaries, forcing petroleum science to engage with geochemistry, thermodynamics, electrochemistry, and microbiology. Together with AI, the low-carbon transition forms the "twin engines" of the paradigm transformation in petroleum science.

This disciplinary roadmap is intended to guide research planning, investment, and talent cultivation for decades to come, while fostering interdisciplinary collaboration on a global scale. The authors emphasized that petroleum science is moving from "resource extraction science" toward "smart integrated energy science"—a comprehensive system encompassing subsurface space utilization (energy storage, waste disposal, carbon sequestration), multi-energy coupling (oil-gas-hydrogen-electricity-heat), carbon cycle management, and intelligent decision-making. This transition demands a systemic reconstruction of knowledge systems, educational models, and research organization.

The authors also noted that any list of scientific grand challenges is inherently open and dynamic; as theoretical breakthroughs and engineering practices advance, old challenges will be solved and new ones will emerge. At this moment, nonetheless, clearly defining what we do not know is an essential step toward the future.

Contact the author:

Li-Zhi Xiao, China University of Petroleum, Beijing, 102249, China, xiaolizhi@cup.edu.cn

Funder:

This work was supported by the China Science and Technology Journal Excellence Action Plan. 

Conflict of interest:

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

See the article:

Xiao, L. Z., Jin, Y., Zhang, L.B., et al., One hundred grand challenges in petroleum science. Petroleum Science, Volume 23, Issue 8, 2026, Pages 4477–4487. https://doi.org/10.1016/j.petsci.2026.06.013.