Tyumen Researchers Improve Forecasting of Oil and Gas Reservoirs
Scientists at Rosneft have introduced a new methodology for forecasting oil and gas reservoir properties. The approach is based on quantitative assessment of seismic forecast reliability using geostatistical methods.
At the Intersection of Geology and Mathematics
The new methodology developed by researchers at Rosneft’s Tyumen Research Institute makes it possible to calculate the probability of successful drilling in percentage terms, estimate the approximate thickness of oil-saturated layers in meters, and model geological conditions for reservoir development with greater accuracy.
The approach increases the accuracy of hydrocarbon prospect forecasts, reduces the risk of inefficient drilling, and improves the data foundation for field development planning. This is critical for effective resource and cost management in the upstream segment, covering exploration, appraisal, drilling, and production of oil and natural gas.
The methodology is built on a combination of geostatistical techniques. Its innovative aspect lies in the quantitative evaluation of seismic forecast reliability, placing the development at the intersection of geology and mathematics. As a result, specialists gain access to rigorously validated information for decision-making.
The contribution made by the researchers has received high-level recognition, including an official commendation from the President of the Russian Federation.
Strong Long-Term Outlook
The development aligns fully with broader strategic objectives. Strengthening technological capabilities is one of the key pillars of Rosneft’s corporate strategy, Rosneft 2030, with innovation positioned as a driver of long-term technological leadership.
On the domestic market, the methodology is expected to see wide adoption across Rosneft’s assets and those of its partners. Experts do not rule out the inclusion of reservoir assessment and drilling planning tools into corporate standards. The next logical step, according to industry specialists, will be the integration of specialized AI-based analysis of geological data.
The export potential of the development is also considered significant. The methodology is particularly relevant for companies operating in complex geological environments, where improved forecast accuracy directly translates into lower operational risk. The approach could also be integrated into international exploration and development projects.
In a Broader Industry Context
At the exploration stage, decision-makers often face a lack of reliable information, while the financial stakes associated with drilling remain high. Exploration drilling decisions are frequently made under conditions of uncertainty.
In 2020, the professional community reported on new applications of geostatistics that enabled more objective assessment of geological probabilities. The approach stood out for its flexibility, allowing experts to work with varying volumes and types of available data. The workflow uses conceptual geological models as inputs and applies geostatistical techniques to generate spatial variability of reservoir properties near prospective drilling targets. As a result, probabilistic reservoir characteristics began to play a more prominent role in evaluation processes.
Last year, the Multi-TransFKAN model entered the market, combining architectural advantages with an enhanced Kolmogorov–Arnold network. The model demonstrates improved performance in handling complex periodic and nonlinear data characteristics compared with earlier solutions.
Parallel efforts are underway at the Gazprom VNIIGAZ Center, the principal scientific organization of Gazprom in the field of technology development. The center focuses on research and methodological support for Gazprom projects across gas production, transportation, underground storage, processing, and utilization. A key priority for VNIIGAZ is the creation of integrated three-dimensional reservoir models and support for geological exploration decisions.
As a result, the role of statistical methods in exploration continues to grow steadily.
