Introduction. Adaptation of Integrated Asset Model (IAM) component models is an important step in the creation of a mathematical model of an oil and gas field. Correct adaptation allows for the generation of production profiles that take into account all infrastructure constraints and bottlenecks in the production system. Through the large-scale integration of development tools, application programming interfaces, integrated modeling tool suites, and real-time data management tools, solutions for the rapid and automated adaptation (auto-adaptation) of oil and gas asset models can be implemented.

Aim. The aim of this study is to describe a methodology for automated history matching of reservoir (material balance model), well, and gathering and transport network models (GTN), as well as experience implementing them in a soft ware product.

Materials and methods. The source materials for this study were drawn from the many years of experience and best practices of subject matter experts collected by the authors, as well as synthetic and existing production model components of the IAM of varying complexity, to formalize a unique method for identifying tuning parameters. To optimize calculations based on these methods, a modular computational and analytical soft ware package was implemented in the Python programming language. Testing of the methods and modules of the package was conducted on facility models of the Company's production entities, including Gazpromneft -Khantos, Gazpromneft -Vostok, and Gazpromneft -Orenburg.

Results. The methods and modules used in this study solve the problem of automated history matching of material balance, well, and GTN models without the need for a subject matter expert to directly participate in the history matching process. The methodology and modules for automatic model matching enable models to be tuned to current actual data with a minimum number of model runs. Auxiliary modules help assess the quality of the history matching process and the final convergence of calculated model indicator values with actual data.

Conclusion. The advantages of the algorithms include: 1) automation of the history matching process; 2) a wide range of supported types and parameters for component model matching; 3) speed of reservoir and GTN models matching after a small number of model runs; 4) scalability of the soft ware package and universality of approaches for a large number of integrated modeling tools (including domestic import-independent soft ware). The acceptable error level of automated matching process for all model types was determined. The results of the calculation and analytical soft ware package allow us to conclude that the proposed solution is effective and demonstrate the importance of automating the history matching process.