The work tells about the experience of development and application of the cognitive analytical system prototype based on the approaches and principles of machine learning for the processing and analysis of statistical and dynamic data collected during the field exploitation. The developed system prototype implements an efficient and modern data-driven approach to field development management, and is an example of a paradigm shift in reservoir modeling.
In the process of creating a prototype of a cognitive analytical system for managing field development, both classical machine learning methods and modified capacitive-resistive models (CRM), INSIM models were used, as well as the approach of identifying a nonlinear dynamic system through recurrent neural networks (RNN). This system provides rapid analysis of a mature field development and to handle such management tasks as optimization of the water flooding, identification of inefficient injection and production, evaluation of the wells interaction, forecast of liquid and oil production at a given time horizon. The system consists of the models that have weak points and limitations, and is not an ultimate decision-making system, but should be used in conjunction with other tools for reservoir engineering. This tool will be an addition to the approaches of the "Digital Field", in which the field development is optimized not only based on the results of the classic physics models, but also on machine learning technologies applied to the entire development history.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.