Abstract
ABSTRACT: The correlation of rock mechanical properties from one well to another across an area of interest poses a classical and ongoing problem in rock mechanics. This work illustrates identification of the mechanical layers/zones in a geothermal reservoir using unsupervised machine learning (ML) techniques. Mechanical stratigraphy was defined using well logs obtained from three wells located at the Utah FORGE geothermal site: 58-32, 16A(78)-32 and 16B(78)-32. The widely accepted unsupervised ML techniques including K-means clustering, Gaussian mixture models, and DBSCAN (density-based spatial clustering of applications with noise) were utilized to generate the rock classes based on similarities/differences in mechanical attributes. The rock mechanical classifications were performed using a combination of parameters including measured log data (compressional and shear wave interval transit times) and augmented features such as Poisson's ratio, and Young's modulus. The performance of ML clustering models were evaluated using Silhouette index (SI) and Davies-Bouldin index (DBI) criteria. The evaluation measures of predicted classification reflected the effectiveness and applicability of the proposed ML approaches to generate mechanical stratigraphy. Evaluation measures SS and DBI represent the good quality and reliability of proposed classification with higher SI, CHI, and lower DBI scores. The best performance for the proposed clustering model was exhibited by K-means algorithm with SI, DBI and CHI scores of 0.86, 0.4, and 79, respectively. The proposed mechanical units cluster models were observed to be consistent with the lithological stratigraphy of the studied wells. This approach is therefore shown to provide efficient and reliable identification of mechanical stratigraphy for FORGE with the capability for application across a wide range of subsurface reservoirs. 1. INTRODUCTION Rocks are formed in different lithostratigraphic units that have a wide range of mechanical characteristics (Boersma et al., 2020). According to Ferrill et al. (2017) and Smart et al. (2014). The mechanical characteristics are often described in terms of stiffness and strength properties, including elastic parameters, tensile strength, and compressive strength (Roche et al., 2013).
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.