A computational algorithm based on an artificial neural network has been developed and implemented for lithological interpretation of well logging data for the Bazhenov Formation. The algorithm estimates the mineral-component composition of rocks. Our studies employ the classification of the Bazhenov Formation lithological types, which is centered on the modern concept of rock-forming mineral and mineraloid components distribution (clay, siliceous, carbonate minerals, and organic matter). Using the developed algorithm for a set of well logging data and taking into account the results of laboratory lithological and geochemical core studies, we constructed models of the content of rock-forming components of the Bazhenov Formation for the central part of the Salym field. The main lithological types of Bazhenov Formation rocks were distinguished: silicites, mudstones, carbonates, and mixtites (mixed siliceous-clay-carbonate rocks), including those enriched in organic matter. The contents of rock-forming components that were calculated with usage of an artificial neural network have a good correlation with the results of detailed lithological and geochemical core studies. Based on the obtained lithological models, we constructed correlation schemes of the Bazhenov Formation, which made it possible to trace the vertical and lateral variability of its mineral-component composition. The average contents of clay, siliceous, carbonate minerals, pyrite, albite, and organic matter have been determined. Significant spatial heterogeneity of the Bazhenov Formation is observed due to the multicomponent composition and complex distribution of various types of rocks that affect its main characteristic features within the local area of the examined field. The obtained results of the performed studies can be useful in research of the structure of the Bazhenov Formation when core materials are unavailable.
References
1. Epov M.I., Glinskikh V.N., Petrov A.M. et al., Frequency dispersion of electrophysical characteristics and resistivity anisotropy of the Bazhenov formation deposits according to resistivity logging data (In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2019, no. 9, pp. 62–64, DOI: 10.24887/0028-2448-2019-9-62-64
2. Kulyapin P.S., Sokolova T.F., Prediction of reservoirs in the section of the Bazhenov formation based on core samples and geophysical studies of wells (In Russ.), Izvestiya Tomskogo politekhnicheskogo universiteta, 2015, V. 326, no. 1, pp. 118 – 133.
3. Cybenko G., Approximation by superpositions of a sigmoidal function, Mathematics of control, signals and systems, 1989, V. 2, no. 4, pp. 303–314.
4. Al-Mudhafar W.J., Integrating well log interpretations for lithofacies classification and permeability modeling through advanced machine learning algorithms, Journal of Petroleum Exploration and Production Technology, 2017, V. 7, no. 4, pp. 1023–1033.
5. Khabarov A.V., Oshnyakov I.O., Aleksandrova I.O. et al., A multidimensional analysis of logs and core as a tool for the petrophysical typing of the Bazhenov-Abalak association (In Russ.), Karotazhnik, 2019, V. 300, no. 6, pp. 86–102.
6. Kontorovich A.E., Yan P.A., Zamiraylova A.G. et al., Classification of rocks of the Bazhenov formation (In Russ.), Geologiya i geofizika, 2016, V. 57, no. 11, pp. 2034–2043.
7. Kingma D.P., Ba J., Adam: A method for stochastic optimization, ArXiv: 1412.6980, 2014.
8. Goodfellow I., Bengio Y., Courville A., Deep learning, Cambridge: MIT press, 2016, V. 1, no. 2, 787 p.
9. Eder V.G., Zamiraylova A.G., Zanin Yu.N., Zhigul'skiy I.A., Lithological structural features of the main types of the Bazhenov formation sections (In Russ.), Geologiya nefti i gaza, 2015, no. 6, pp. 96–106.
