Home Petroleum Science Article
PDF (2.6 MB)
Cite
EndNote(RIS) BibTeX
Collect
Submit Manuscript
Show Outline
Outline
Abstract
Keywords
Electronic Supplementary Material
References
Show full outline
Hide outline
Original Paper | Open Access

Effect of preprocessing on performances of machine learning-based mineral composition analysis on gas hydrate sediments, Ulleung Basin, East Sea

Hongkeun Jina,b,1, Ju Young Parka,f,1, Sun Young Parkc, Byeong-Kook Sonc, Baehyun Mind,e, Kyungbook Leea,f()
Department of Geoenvironmental Sciences, Kongju National University, Gongju-si, Chungcheongnam-Do, 32588, Republic of Korea
Tunnel & Underground Facility Division, ESCO Consultant & Engineers Company, Simin-Daero, Dongan-Gu, Anyang-Si, Gyeonggi-Do, 14057, Republic of Korea
Petroleum and Marine Research Division, Korea Institute of Geoscience and Mineral Resources, 124, Gwahak-ro, Yuseong-gu, Daejeon, 34132, Republic of Korea
Department of Climate and Energy Systems Engineering, Ewha Womans University, 52 Ewhayeodae-gil, Seodaemun-gu, Seoul, 03760, Republic of Korea
Center for Climate/Environment Change Prediction Research, Ewha Womans University, 52 Ewhayeodae-gil, Seodaemun-gu, Seoul, 03760, Republic of Korea
Yellow Sea Institute of Geoenvironmental Sciences, Gongju-si, Chungcheongnam-Do, 32588, Republic of Korea

1 These authors contributed equally to this work.

Edited by Teng Zhu

Show Author Information

Abstract

Gas hydrate (GH) is an unconventional resource estimated at 1000–120,000 trillion m3 worldwide. Research on GH is ongoing to determine its geological and flow characteristics for commercial production. After two large-scale drilling expeditions to study the GH-bearing zone in the Ulleung Basin, the mineral composition of 488 sediment samples was analyzed using X-ray diffraction (XRD). Because the analysis is costly and dependent on experts, a machine learning model was developed to predict the mineral composition using XRD intensity profiles as input data. However, the model’s performance was limited because of improper preprocessing of the intensity profile. Because preprocessing was applied to each feature, the intensity trend was not preserved even though this factor is the most important when analyzing mineral composition. In this study, the profile was preprocessed for each sample using min-max scaling because relative intensity is critical for mineral analysis. For 49 test data among the 488 data, the convolutional neural network (CNN) model improved the average absolute error and coefficient of determination by 41% and 46%, respectively, than those of CNN model with feature-based preprocessing. This study confirms that combining preprocessing for each sample with CNN is the most efficient approach for analyzing XRD data. The developed model can be used for the compositional analysis of sediment samples from the Ulleung Basin and the Korea Plateau. In addition, the overall procedure can be applied to any XRD data of sediments worldwide.

Keywords

Sample-based preprocessingX-ray diffraction (XRD)Machine learningMineral compositionGas hydrate (GH)Ulleung basin

Electronic Supplementary Material

Download File(s)
petsci-22-1-151_ESM.pdf (582.9 KB)

References

 

Ainane, A., Taleb, M., El-Hajjaji, F., Hammouti, B., Chatouani, A., Ainane, T., 2021. Study of dependence between two types of most abundant natural clays in Bejaad province (Central Morocco) using a statistical approach. Moroc. J. Chem. 9, 210–220. https://doi.org/10.48317/IMIST.PRSM/morjchem-v9i2.22438.

Crossref Google Scholar
 

Domínguez-Olmedo, J.L., Toscano, M., Mata, J., 2020. Application of classification trees for improving optical identification of common opaque minerals. Comput. Geosci. 140, 104480. https://doi.org/10.1016/j.cageo.2020.104480.

Crossref Google Scholar
 

Dong, R., Zhao, Y., Song, Y., Fu, N., Omee, S.S., Dey, S., Li, Q., Wei, L., Hu, J., 2022. DeepXRD, a deep learning model for predicting XRD spectrum from material composition. ACS Appl. Mater. Interfaces 14, 40102–40115. https://doi.org/10.48550/arXiv.2203.14326.

Crossref Google Scholar
 

Gers, F.A., Schmidhuber, J., Cummins, F., 1999. Learning to forget: continual prediction with LSTM. Ninth Int. Conf. Artific. Neural Networks ICANN 99 (2), 850–855. https://doi.org/10.1049/cp:19991218.

Crossref Google Scholar
 

Ji, M.S., Kwon, S.Y., Kim, M., Kim, S.I., Min, B.H., 2022. Generation of synthetic compressional wave velocity based on deep learning: a case study of Ulleung Basin gas hydrate in the Republic of Korea. Appl. Sci. 12, 8775. https://doi.org/10.3390/app12178775.

Crossref Google Scholar
 

Kim, D.K., Choi, J.H., Kim, D.W., Byun, J.M., 2020. Predicting mineralogy by integrating core and well log data using a deep neural network. J. Petrol. Sci. Eng. 195, 107838. https://doi.org/10.1016/j.petrol.2020.107838.

Crossref Google Scholar
 

Kvenvolden, K.A., Ginsburg, G.D., Soloviev, V.A., 1993. Worldwide distribution of subaquatic gas hydrates. Geo Mar. Lett. 13, 32–40. https://doi.org/10.1007/BF01204390.

Crossref Google Scholar
 

Lee, J.-W., Park, W.B., Lee, J.H., Singh, S.P., Sohn, K.-S., 2020. A deep-learning technique for phase identification in multiphase inorganic compounds using synthetic XRD powder patterns. Nat. Commun. 11, 704. https://doi.org/10.1038/s41467-020-14512-9.

Crossref Google Scholar
 

Li, Y., Liu, L., Jin, Y., Wu, N., 2021. Characterization and development of natural gas hydrate in marine clayey−silt reservoirs: a review and discussion. Adv. Geo-Energy Res. 5, 75–86. https://doi.org/10.46690/ager.2021.01.08.

Crossref Google Scholar
 
McLennan, S.M., Hemming, S., McDaniel, D.K., Hanson, G.N., 1993. Geochemical approaches to sedimentation, provenance and tectonics. In: Johnsson, M.J., Basu, A. (Eds.), Processes Controlling the Composition of Clastic Sediments, vol. 284. Spec. Pap. Geol. Soc. Am., pp. 21–40. https://doi.org/10.1130/SPE284-p21
Crossref
 

Okada, N., Maekawa, Y., Owada, N., Haga, K., Shibayama, A., Kawamura, Y., 2020. Automated identification of mineral types and grain size using hyperspectral imaging and deep learning for mineral processing. Minerals 10, 809. https://doi.org/10.3390/min10090809.

Crossref Google Scholar
 

Park, S.Y., Son, B.-K., Choi, J.Y., Jin, H.K., Lee, K.B., 2022. Application of machine learning to quantification of mineral composition on gas hydrate-bearing sediments, Ulleung Basin, Korea. J. Petrol. Sci. Eng. 209, 109840. https://doi.org/10.1016/j.petrol.2021.109840.

Crossref Google Scholar
 

Rietveld, H.M., 1969. A profile refinement method for nuclear and magnetic structures. J. Appl. Crystallogr. 2, 65–71. https://doi.org/10.1107/S0021889869006558.

Crossref Google Scholar
 

Ryu, B.-J., Collett, T.S., Riedel, M., Kim, G.Y., Chun, J.-H., Bahk, J.-J., Lee, J.Y., Kim, J.-H., Yoo, D.-G., 2013. Scientific results of the second gas hydrate drilling expedition in the Ulleung Basin (UBGH2). Mar. Petrol. Geol. 47, 1–20. https://doi.org/10.1016/j.marpetgeo.2013.07.007.

Crossref Google Scholar
 

Schuetzke, J., Benedix, A., Mikut, R., Reischl, M., 2021. Enhancing deep-learning training for phase identification in powder X-ray diffractograms. IUCrJ 8, 408–420. https://doi.org/10.1107/S2052252521002402.

Crossref Google Scholar
 

Shahnazar, S., Hasan, N., 2014. Gas hydrate formation condition: review on experimental and modeling approaches. Fluid Phase Equil. 379, 72–85. https://doi.org/10.1016/j.fluid.2014.07.012.

Crossref Google Scholar
 

Sietronics, 1996. SIROQUANT: A Quantitative XRD Software. Sietronics Pty Limited, Belconnen ACT, Australia.

 

Taylor, J.C., 1991. Computer programs for standardless quantitative analysis of minerals using the full powder diffraction profile. Powder Diffr. 6, 2–9. https://doi.org/10.1017/S0885715600016778.

Crossref Google Scholar
 

Taylor, J.C., Matulis, C.E., 1994. A new method of rietveld clay analysis. Part Ⅰ. Use of a universal measured standard profile for rietveld quantification of montmorillonite. Power Diffract. 9, 119–123. https://doi.org/10.1017/S0885715600014093.

Crossref Google Scholar
 

Wang, R., Shi, W., Xie, X., Zhang, W., Qin, S., Liu, K., Busbey, A.B., 2020. Clay mineral content, type, and their effects on pore throat structure and reservoir properties: Insight from the Permian tight sandstones in the Hangjinqi area, north Ordos Basin, China. Mar. Petrol. Geol. 115, 104281. https://doi.org/10.1016/j.marpetgeo.2020.104281.

Crossref Google Scholar
 

Zeng, X., Xiao, Y., Ji, X., Wang, G., 2021. Mineral identification based on deep learning that combines image and mohs hardness. Minerals 11, 506. https://doi.org/10.3390/min11050506.

Crossref Google Scholar
 

Zhang, F., An, M., Zhang, L., Fang, Y., Elsworth, D., 2019. The role of mineral composition on the frictional and stability properties of powdered reservoir rocks. J. Geophys. Res. Solid Earth 124, 1480–1497. https://doi.org/10.1029/2018JB016174.

Crossref Google Scholar
Petroleum Science
Volume 22 Issue 1,
January 2025
Pages 151-162
DOI: 10.1016/j.petsci.2024.11.012
Cite this article:
Jin H, Park JY, Park SY, et al. Effect of preprocessing on performances of machine learning-based mineral composition analysis on gas hydrate sediments, Ulleung Basin, East Sea. Petroleum Science, 2025, 22(1): 151-162. https://doi.org/10.1016/j.petsci.2024.11.012
Metrics & Citations  
Article History
Copyright
Rights and Permissions
Return