TY - JOUR
T1 - Interpreting the Subsurface Lithofacies at High Lithological Resolution by Integrating Information From Well-Log Data and Rock-Core Digital Images
AU - Jeong, Jina
AU - Park, Eungyu
AU - Emelyanova, Irina
AU - Pervukhina, Marina
AU - Esteban, Lionel
AU - Yun, Seong Taek
N1 - Funding Information:
This work was supported by the Korea Environmental Industry and Technology Institute (KEITI) (project title: Development and Field Verification of Environmental Risk Estimation System for CO2 Leakage; Project 2018001810004) and by the Commonwealth Scientific and Industrial Research Organization (CSIRO), Australia. The associated source codes (MATLAB and TensorFlow), data, the mineralogy data from the HyLogger system, and output files that support the findings of this study are available online (at https://doi.org/10.5281/zenodo.3603553). The well logs and rock-core digital images can be also acquired from the National Offshore Petroleum Information Management System (NOPIMS; at the following address: https://nopims.dmp.wa.gov.au/Nopims/Search/WellDetails#). The authors thank Lena Hancock from the Geoscience Australia and Geological Survey of Western Australia Perth Core Library (Carlisle, Western Australia) to perform the HyLogger measurements on the Satyr 5 core sections. National Offshore Petroleum Information Management System (NOPIMS) was used to obtain the Satyr 5 log data and well completion reports. Additional information is available from Jina Jeong (contact email: jeong.j@knu.ac.kr) on request.
Publisher Copyright:
©2020. American Geophysical Union. All Rights Reserved.
PY - 2020/2/1
Y1 - 2020/2/1
N2 - Spectral facies interpretation and classification methods have been proposed to improve the sophistication of interpretation of the subsurface heterogeneity. In the spectral facies interpretations, the intensity values of the RGB spectrum and the local entropy from rock-core digital images are used, and the results are compared to conventional electrofacies and expert petrophysical interpretations. During the classification, a practically applicable model that identifies the more detailed types of lithofacies is constructed by using a multilayer neural network model, with the interpreted spectral facies and well-log data from the corresponding depths used as response and explanatory variables, respectively. Core digital images and five types of well-log data from the Satyr 5 well in Western Australia are applied for the actual implementation. Through comparative interpretations, three spectral facies are identified as separable lithofacies (i.e., shale, shaly-sandstone, and sandstone lithofacies), which is supported by detailed HyLogger mineralogy along the tested cores. On the other hand, two electrofacies (i.e., shale-dominant and sand-dominant facies) are identified by a conventional method. In the classification based on the spectral facies, the trained multilayer neural network model showed high prediction accuracy for all the lithofacies. Based on these observations, it is confirmed that more precise lithofacies interpretation and classification can be conducted with the developed methods. The developed methods have the potential to improve subsurface characterization when high lithological resolution is essential.
AB - Spectral facies interpretation and classification methods have been proposed to improve the sophistication of interpretation of the subsurface heterogeneity. In the spectral facies interpretations, the intensity values of the RGB spectrum and the local entropy from rock-core digital images are used, and the results are compared to conventional electrofacies and expert petrophysical interpretations. During the classification, a practically applicable model that identifies the more detailed types of lithofacies is constructed by using a multilayer neural network model, with the interpreted spectral facies and well-log data from the corresponding depths used as response and explanatory variables, respectively. Core digital images and five types of well-log data from the Satyr 5 well in Western Australia are applied for the actual implementation. Through comparative interpretations, three spectral facies are identified as separable lithofacies (i.e., shale, shaly-sandstone, and sandstone lithofacies), which is supported by detailed HyLogger mineralogy along the tested cores. On the other hand, two electrofacies (i.e., shale-dominant and sand-dominant facies) are identified by a conventional method. In the classification based on the spectral facies, the trained multilayer neural network model showed high prediction accuracy for all the lithofacies. Based on these observations, it is confirmed that more precise lithofacies interpretation and classification can be conducted with the developed methods. The developed methods have the potential to improve subsurface characterization when high lithological resolution is essential.
KW - Gaussian mixture model
KW - lithofacies interpretation
KW - multilayer neural network
KW - rock-core digital images
KW - spectral facies
KW - well-log data
UR - http://www.scopus.com/inward/record.url?scp=85081035369&partnerID=8YFLogxK
U2 - 10.1029/2019JB018204
DO - 10.1029/2019JB018204
M3 - Article
AN - SCOPUS:85081035369
VL - 125
JO - Journal of Geophysical Research: Solid Earth
JF - Journal of Geophysical Research: Solid Earth
SN - 2169-9313
IS - 2
M1 - e2019JB018204
ER -