Abstract

LEAL F, Jorge A; OCHOA G, Luis H  and  SARMIENTO P, Gustavo A. Content of Total Organic Carbon Using Random Forest, Borehole Imaging, and Fractal Analysis: A Methodology Applied in the Cretaceous La Luna Formation, South America. Geofís. Intl [online]. 2022, vol.61, n.4, pp.301-323.  Epub Nov 18, 2022. ISSN 2954-436X.  https://doi.org/10.22201/igeof.00l67l69p.2022.6l.4.2113.

This research presents an alternative approach to computing the content of total organic carbon using wireline logs and machine learning techniques. Specifically, borehole resistivity imaging, its average resistivity, and gamma rays log are employed to train a regression model. The methodology was applied in La Luna Formation, which has been reported as one of the principal source rocks for Colombia and western Venezuela. This work aims to teach a machine how to recognize patterns between fractal features in borehole images and their content of total organic carbon. Implemented machine learning is based on ensemble learning techniques, in this case, an ensemble of decision trees known as random forest. The working data set totalizes 960 wireline log measurements, randomly split into 80% for training and 20% for validation. The outcome is equivalent to the curve obtained using a semi-log regression of organic carbon measured in core against density log values. The accuracy of this method is high enough to be considered during petrophysics evaluations, showing a root-mean-square error of 0.44% and Pearson's correlation coefficient of 0.88. The methodology depends on image quality, and anomalies in these data increase the error. The generated model must be recalibrated for other formations, for horizontal and deviated wells, and when logging while drilling imaging is employed.

Keywords : La Luna Formation; total organic carbon; borehole resistivity imaging; random forest; fractal analysis and unconventional reservoirs.

        · abstract in Spanish     · text in English     · English ( pdf )