The strength of Machine Learning-based technology is its ability to deal with massive amounts of data from various sources, learn from them, and transform them to deliver comprehensible information to the end user. To obtain a more realistic behavior of a reservoir, geoscientists try to achieve accurate lithofacies distribution mapping. Accordingly, predicting rock type quality distribution with seismic data will enable geoscientists and engineers to better understand depositional processes for optimizing landing target selection and well design. In this paper, we discuss the application of a Machine Learning method to resolve the seismic-scale mapping of reservoir facies heterogeneities in the unconventional shale oil-rich TOC reservoir of the lower section of the Vaca Muerta Formation.
Machine Learning and Neural Networks are now commonly used in both the traditional oil and gas industry and in activities related to energy transition. The strength of Machine Learning-based technology is its ability to deal with a massive amount of data from various sources, learn from them, and transform them for delivering understandable information to the end user.
Traditionally, the more data we acquire, and the more we create and interpret them in separate ways, the more challenging it is to integrate them into a single framework to improve our knowledge of the subsurface.
To evaluate the quality of a reservoir and obtain a more realistic measure of its behavior, geoscientists try to achieve accurate lithofacies distribution mapping. Accordingly, predicting rock type quality distribution with seismic data will enable geoscientists and engineers to better understand depositional processes for optimizing landing target selection and well design.
To perform the seismic-scale mapping of reservoir facies heterogeneities, we discuss the application of a Machine Learning method to the unconventional shale oil-rich TOC reservoirs of the lower section of the Vaca Muerta Formation. The objective of the asset team is to better understand the shale oil reservoir properties and their associated facies distribution, in order to apply them in geocellular modeling. The key technology is provided by Neural Networks.
