Abstract
Subsurface interpretation and modeling are crucial to the success of reservoir exploration and production, which often involves integrating multiple types of subsurface data and accomplishing a series of subtasks consecutively and/or in parallel. Those tasks include data processing and conditioning, structural mapping, property modeling, and others. With the emergence of machine learning (ML), each component of the subsurface modeling and interpretation workflow now has elements of ML automation incorporated into individual steps. However, there have been few attempts to generate an integrated end-to-end workflow that incorporates ML at every stage. This paper presents an end-to-end solution for subsurface modeling and interpretation that is powered by multiple convolutional neural networks (CNNs). Its performance is demonstrated on the Groningen gas field. The workflow can be subdivided into four parts, with the initial phase focusing on data preconditioning, particularly log quality control and reconstruction. Phase two focuses on seismic property estimation (i.e., relative geologic time) and structural feature identification and extraction (i.e., fault and salt). Phase three combines seismic feature extraction, log data, derived seismic properties, and interpretations to generate a 3D rock property model containing modeled gamma ray, density, velocity, and porosity properties. The final component is to validate the estimated subsurface properties by CNN-based seismic image simulation. The results demonstrated in the Groningen field show high accuracy, strong lateral consistency, and a good match with the dominant subsurface features documented in this area, including the Zechstein salt, the complex fault system, and the Rotliegend reservoir model. We conclude that this end-to-end workflow can be readily applied to other fields to provide a one-click solution for efficient subsurface modeling and interpretation.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.