Machine-Learning-Assisted Field Development Opportunity Identification Through Streamlined Geological and Engineering Workflows

Summary As an essential aspect of the field development planning process, the identification of actionable field development opportunities such as recompletions and sweet spots remains a high priority for asset teams. Field development planning usually involves extensive geological understanding, deep reservoir model analysis, historical production performance tracking, multi-discipline collaboration, etc. Due to the complexity of the workflow, traditional processes that involve data gathering, vetting, and analyzing are typically labor-intensive and time-consuming. In the workflow presented in this research, standard time-consuming workflows from both geological and engineering aspects have been streamlined and automated to enable fast decision-making in time-sensitive field development and acquisition plans. The presented workflow automates steps asset teams typically perform by applying advanced algorithms to multi-disciplinary datasets, enabling teams to rapidly review options in future development planning. It also integrates multiple disparate data sources and opens new cross-functional workflows. The typical standardized steps include by-passed pay zone identification, drainage analysis, geo-engineering attribute mapping, production forecast, risk quantification, etc. The application of machine learning algorithms also greatly contributes to multiple processes such as data imputation, well log interpretation, production forecast, etc. The final target is to generate a comprehensive list of opportunities for recompletions, sweet spots, and horizontal wells.