Ensemble Deep Learning-Based Porosity Inversion From Seismic Attributes
Underground porosity is important in many earth sciences and engineering fields, including hydrocarbon reservoir characterization and geothermal energy production. Popular methods largely rely on the analysis of lithological core data, well logs, and seismic inversion methods. While these methods ar...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10025713/ |
| _version_ | 1797938792138539008 |
|---|---|
| author | Jianguo Song Munezero Ntibahanana Moise Luemba Keto Tondozi Gloire Imani |
| author_facet | Jianguo Song Munezero Ntibahanana Moise Luemba Keto Tondozi Gloire Imani |
| author_sort | Jianguo Song |
| collection | DOAJ |
| description | Underground porosity is important in many earth sciences and engineering fields, including hydrocarbon reservoir characterization and geothermal energy production. Popular methods largely rely on the analysis of lithological core data, well logs, and seismic inversion methods. While these methods are reliable, they are also time-consuming, expensive, and difficult to implement. In addition, seismic inversion has nonlinearity, data dimensionality, and non-uniqueness issues. However, deep learning (DL) can provide a more flexible, efficient, and accurate capability by mapping directly from seismic attributes to underground porosity. Therefore, we trained several DL models with different optimization functions. In the training steps, we labelled every seismic attribute data point with its corresponding porosity derived from the well-logs. In contrast to popular ensemble techniques, we proposed a weighted prediction approach based on the strengths of each model. Testing results showed a coefficient of determination (<inline-formula> <tex-math notation="LaTeX">$\text{R}^{2}$ </tex-math></inline-formula>) of 0.94345 and a Pearson’s correlation coefficient of 0.9725 between the actual model and the model of the proposed approach, versus 0.9681 and 0.9716 for the best single and popular ensemble models, respectively. Further, we tested the effectiveness of our method using real seismic data from the North Sea. With a Pearson’s correlation value of 0.9743, the inverted model ranges from 27 to 35%, compared to the reference model, which has an overall range of 20 to 33%. These results provide insights into the potential of the proposed method and its applicability to any other seismic volume to determine spatially varying underground porosity from seismic attributes directly. |
| first_indexed | 2024-04-10T19:06:22Z |
| format | Article |
| id | doaj.art-183c63f1cac746d0a08ba12c0a512b34 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-10T19:06:22Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-183c63f1cac746d0a08ba12c0a512b342023-01-31T00:00:17ZengIEEEIEEE Access2169-35362023-01-01118761877210.1109/ACCESS.2023.323968810025713Ensemble Deep Learning-Based Porosity Inversion From Seismic AttributesJianguo Song0https://orcid.org/0000-0003-0302-7201Munezero Ntibahanana1https://orcid.org/0000-0002-2082-0853Moise Luemba2Keto Tondozi3Gloire Imani4https://orcid.org/0000-0002-5126-8491School of Geosciences, China University of Petroleum (East China), Qingdao, Shandong, ChinaSchool of Geosciences, China University of Petroleum (East China), Qingdao, Shandong, ChinaSchool of Geosciences, China University of Petroleum (East China), Qingdao, Shandong, ChinaFaculté de Sciences, Université de Kinshasa, Kinshasa, DR, CongoSchool of Petroleum Engineering, China University of Petroleum (East China), Qingdao, Shandong, ChinaUnderground porosity is important in many earth sciences and engineering fields, including hydrocarbon reservoir characterization and geothermal energy production. Popular methods largely rely on the analysis of lithological core data, well logs, and seismic inversion methods. While these methods are reliable, they are also time-consuming, expensive, and difficult to implement. In addition, seismic inversion has nonlinearity, data dimensionality, and non-uniqueness issues. However, deep learning (DL) can provide a more flexible, efficient, and accurate capability by mapping directly from seismic attributes to underground porosity. Therefore, we trained several DL models with different optimization functions. In the training steps, we labelled every seismic attribute data point with its corresponding porosity derived from the well-logs. In contrast to popular ensemble techniques, we proposed a weighted prediction approach based on the strengths of each model. Testing results showed a coefficient of determination (<inline-formula> <tex-math notation="LaTeX">$\text{R}^{2}$ </tex-math></inline-formula>) of 0.94345 and a Pearson’s correlation coefficient of 0.9725 between the actual model and the model of the proposed approach, versus 0.9681 and 0.9716 for the best single and popular ensemble models, respectively. Further, we tested the effectiveness of our method using real seismic data from the North Sea. With a Pearson’s correlation value of 0.9743, the inverted model ranges from 27 to 35%, compared to the reference model, which has an overall range of 20 to 33%. These results provide insights into the potential of the proposed method and its applicability to any other seismic volume to determine spatially varying underground porosity from seismic attributes directly.https://ieeexplore.ieee.org/document/10025713/Deep neural networksensemble methodsseismic inversionreservoir properties |
| spellingShingle | Jianguo Song Munezero Ntibahanana Moise Luemba Keto Tondozi Gloire Imani Ensemble Deep Learning-Based Porosity Inversion From Seismic Attributes IEEE Access Deep neural networks ensemble methods seismic inversion reservoir properties |
| title | Ensemble Deep Learning-Based Porosity Inversion From Seismic Attributes |
| title_full | Ensemble Deep Learning-Based Porosity Inversion From Seismic Attributes |
| title_fullStr | Ensemble Deep Learning-Based Porosity Inversion From Seismic Attributes |
| title_full_unstemmed | Ensemble Deep Learning-Based Porosity Inversion From Seismic Attributes |
| title_short | Ensemble Deep Learning-Based Porosity Inversion From Seismic Attributes |
| title_sort | ensemble deep learning based porosity inversion from seismic attributes |
| topic | Deep neural networks ensemble methods seismic inversion reservoir properties |
| url | https://ieeexplore.ieee.org/document/10025713/ |
| work_keys_str_mv | AT jianguosong ensembledeeplearningbasedporosityinversionfromseismicattributes AT munezerontibahanana ensembledeeplearningbasedporosityinversionfromseismicattributes AT moiseluemba ensembledeeplearningbasedporosityinversionfromseismicattributes AT ketotondozi ensembledeeplearningbasedporosityinversionfromseismicattributes AT gloireimani ensembledeeplearningbasedporosityinversionfromseismicattributes |