Carbonate deposits are important and promising objects of the study of hydrocarbon resources, but the features of their geological structure and significant variability of rock properties create certain difficulties in identifying productive intervals. In the traditional approach to interpretation of well logging data, productive intervals are identified based on petrophysical modelling and the use of data from an extended set of well logging data. Petrophysical modelling enables to determine the main characteristics of carbonate rocks based on the results of laboratory studies on core samples. However, in some cases (for example, when there are no data from special logging methods and core studies in the well), the quality of reservoir property prediction is significantly reduced to the point of impossibility of using models in practice. To solve these problems at the studied fields, it is necessary to apply new methods for interpreting well logging data. The paper proposes an approach to modelling data of acoustic and density logging based on the use of artificial intelligence methods in an automated multi-well mode. This will automate the process of interpreting geophysical information of wells and increase the efficiency of petrophysical modelling of the carbonate reservoirs. The obtained research results open up prospects for further application of artificial intelligence in geophysics in general.
References
1. Serbaeva A.R., Kachkaeva E.A., Suleymanov D.D., Amineva G.R., The influence of abnormal radioactivity on the assessment of geological reserves in carbonate reservoirs of the Artinskian age (In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2023, no. 11, pp. 73–77, DOI: http://doi.org/10.24887/0028-2448-2023-11-73-77
2. Privalova O.R., Gadeleva D.D., Minigalieva G.I. et al., Well logging interpretation for Kashir and Podolsk deposits using neural networks (In Russ.), Neftegazovoe delo, 2021, V. 19, no. 1, pp. 69–76, DOI: http://doi.org/10.17122/ngdelo-2021-1-69-76
3. Plas D.V., Python dlya slozhnykh zadach: nauka o dannykh i mashinnoe obuchenie (Python for complex tasks: Data science and machine learning), St. Petersburg, Piter Publ., 2018,. – 576 s. – ISBN 978-5-496-03068-7.
4. Kazaryan A.A., Lubyanskaya E.A., Approaches to automating the search for intervals of a destroyed wellbore when building an interpretive model based on well logging data (In Russ.), Proceedings of the XXIV Ural Youth Scientific School on Geophysics, 2023, pp. 89–93.
5. Kazaryan A.A., Lubyanskaya E.A., Markov A.V., Reservoir identification using machine learning algorithms (In Russ.), Proceedings of international scientific and practical conference “GeoSochi-2024. Novye idei i tekhnologii razvedochnoy i promyslovoy geofiziki” (GeoSochi-2024. New ideas and technologies of exploration and production geophysics), 2024, pp. 110–113.
6. Kuznetsova I.O., Malyutov D.A., The principle of operation and architecture of neural networks (In Russ.), Collected papers “ Evraziyskaya integratsiya: sovremennye trendy i perspektivnye napravleniya”, 2024, V. 7, pp. 106–111, DOI: http://doi.org/10.24412/cl-37031-2024-2-106-111
7. Basyrov M.A., Sergeychev A.V., Latypov I.D. et al., Application of machine learning methods for the petrophysical interpretation of complex geological section
(In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2024, no. 3, pp. 20-25, DOI: http://doi.org/10.24887/0028-2448-2024-3-20-25
