Machine learning-driven well placement
Challenge
- High uncertainty not captured in infill drilling workflows.
- Lack of automated and rapid assessment workflows.
- Absence of cross-domain integration to evaluate infill well candidate performance.
Solution
This Petrel™ subsurface software plug-in is designed to optimize infill well placement under subsurface uncertainty by automating the identification of hotspots and the design of well trajectories.
- Improved decision making: This solution enables the integration of multiple data sources to build an integrated well placement target map and evaluates candidate well performance within deterministic and probabilistic reservoir modeling workflows, enhancing decision-making for field development planning.
- Enhanced accuracy: By leveraging machine learning, it improves reliability and reduces risks, allowing users to account for uncertainties while adhering to surface constraints and existing drilling centers to enhance the feasibility of the wells for rapid implementation.
Results
1-2 years extension of the oil production plateau.
Automation leads to 80% efficiency in the well optimization task.
Significant savings by reducing well count by more than 15 wells.
