Vol. 337 No. 10 (2026)
DOI https://doi.org/10.18799/24131830/2026/10/5185
Real-time prediction of areal distribution of filtration parameters of oil reservoirs based on machine learning approaches and well-tests
Relevance. Permeability of reservoirs is one of the key parameters influencing the efficiency of oil and gas field development. Most of the currently used methods for determining permeability have several drawbacks, including the need for costly studies. The primary method for assessing the filtration properties of the remote zone of the reservoir is the interpretation of hydrodynamic tests, which are characterized by the requirement for prolonged well shut-ins and the inability to conduct tests simultaneously across the entire well stock. Therefore, an important and relevant task is the creation of a service for evaluating the remote zone permeability of all wells at any given time, which would enable more informed and prompt decision-making for field development management. Aim. To develop an innovative approach for modeling the two-dimensional spatial distribution of filtration parameters in oil field reservoirs. Methods. Interpretation of well testing using specialized software, multi-stage statistical modeling, machine learning methods. Results. The authors have implemented the system for evaluating the permeability of the reservoir remote zone, encompassing sequential stages – from the collection of initial data to the development of predictive models using machine learning tools. The advantages of the developed algorithms include the ability to determine the filtration properties of the reservoir based on hydrodynamic tests simultaneously across the entire well stock. Two models were trained: in the first model, the predicted parameter is the remote zone permeability obtained from the interpretation of hydrodynamic test data. In the second model, the prediction focuses on bottomhole pressure during the pressure recovery phase, modeling a synthetic pressure recovery curve for subsequent interpretation of the remote zone permeability. The developed model (based on the Random Forest method) is based on reliable interpretation values of well testing (in all studies, a section of flat-radial filtration is identified). The coefficient of determination (r) when comparing actual and predicted bottomhole pressures during recovery was 0.963, with a root mean square error of 0.977. The two-dimensional spatial distribution of reservoir filtration properties was constructed for one of the oil fields in the Perm region using spherical kriging.
For citation: Soromotin A.V., Martyushev D.A. Real-time prediction of areal distribution of filtration parameters of oil reservoirs based on machine learning approaches and well-tests. Bulletin of the Tomsk Polytechnic University. Geo Assets Engineering, 2026, vol. 337, no. 10, pp. 65–77. http://doi.org/10.18799/24131830/2026/10/5185
Keywords:
machine learning, regressors, permeability, well tests, pressure recovery curve
References:
1. Iktisanov V.A. Determination of reservoir filtration parameters and rheological properties of fluids during oil field development. Dr. Diss. Bugulma, 2002. 277 p. (In Russ.)
2. Lysenko V.D. Permeability does not depend on porosity. Hydrodynamic permeability should be used in mathematical modeling. Geology, Geophysics and Development of Oil and Gas Fields, 2009, no. 1, pp. 3–10. (In Russ.)
3. Soromotin A.V., Martyushev D.A., Pereira J.L.J. On the application of machine learning algorithms in predicting the permeability of oil reservoirs. Artificial Intelligence in Geosciences, 2025, vol. 6 (2), Article 100126. DOI: https://doi.org/10.1016/j.aiig.2025.100126
4. Byrne M.T., McPhee C.A. The extinction of skin. SPE International Conference and Exhibition on Formation Damage Contro, 2012. DOI: https://doi.org/10.2118/151807-MS.
5. Al-Obaidi S.H., Khalaf F.H., Alwan H.H. Performance analysis of hydrocarbon wells based on the skin zone. Techniсum, 2021, vol. 3, no. 4, pp. 50–56.
6. Belhouchet H.E., Benzagouta M.S., Dobbi A., Alquraishi A., Duplay J. A new empirical model for enhancing well log permeability prediction, using nonlinear regression method: case study from Hassi-Berkine oil field reservoir – Algeria. Journal of King Sau University – Engineering Sciences, 2021, vol. 33, no. 2, pp. 136–145. DOI: https://doi.org/10.1016/j.jksues.2020.04.008
7. Wang X. Improved permeability prediction based on the feature engineering of petrophysics and fuzzy logic analysis in low porosity–permeability reservoir. Journal of Petroleum Exploration and Production Technology, 2019, vol. 9, pp. 869–887. DOI: https://doi.org/10.1007/S13202-018-0556-Y
8. Mirzaei-Paiaman A., Asadolahpour S.R., Saboorian-Jooybari H., Chen Z., Ostadhassan M. A new framework for selection of representative samples for special core analysis. Petroleum Research, 2020, vol. 5, pp. 210–226. DOI: https://doi.org/10.1016/j.ptlrs.2020.06.003
9. Mirzaei-Paiaman A., Saboorian-Jooybari H., Pourafshary P. Improved method to identify hydraulic flow units for reservoir characterization. Energy Technology, 2015, vol. 3, no. 7, pp. 726–733. DOI: https://doi.org/10.1002/ente.201500010
10. Mohammadlou M.H.H., Mørk M.B.B. Integrated permeability analysis in tight and brecciated carbonate reservoir. SPE Reservoir Evaluation & Engineering, 2012, vol. 15, no. 06, pp. 624–635. DOI: https://doi.org/10.2118/131593-pa
11. Zhong Z., Carr T.R. Application of a new hybrid particle swarm optimization-mixed kernels function based support vector machine model for reservoir porosity prediction: a case study in Jacksonburg-Stringtown oil field. Interpretation, 2019, vol. 7, no. 1, T97–T112. DOI: https://doi.org/10.1190/int-2018-0093.1
12. Bennis M., Torres-Verdín C. Automatic multiwell assessment of flow-related petrophysical properties of tight gas sandstones based on the physics of mud-filtrate invasion. SPE Reservoir Evaluation & Engineering, 2023, vol. 26, no. 03, pp. 543–564. DOI: https://doi.org/10.2118/214668-PA
13. Ishola O., Vilcáez J. Machine learning modeling of permeability in 3D heterogeneous porous media using a novel stochastic pore-scale simulation approach. Fuel, 2022, vol. 321, article 124044. DOI: https://doi.org/10.1016/j.fuel.2022.124044
14. Singh M., Makarychev G., Mustapha H. Machine learning assisted petrophysical logs quality control, editing and reconstruction. Abu Dhabi International Petroleum Exhibition and Conference, 2022. DOI: https://doi.org/10.2118/202977-MS
15. Kang Q., Li K., Fu J., Liu Y. Hybrid LBM and machine learning algorithms for permeability prediction of porous media: a comparative study. Computers and Geotechnics, 2024, vol. 168, article 106163. DOI: https://doi.org/10.1016/j.compgeo.2024.106163
16. Jirjees A.Y. Influences of uncertainty in well log petrophysics and fluid properties on well test interpretation: an application in West Al Qurna Oil Field, South Iraq. Egyptian Journal of Petroleum, 2019, vol. 28, no. 4, pp. 383–392. DOI: https://doi.org/10.1016/j.ejpe.2019.08.005
17. Elesin A.V., Kadyrova A.Sh., Nikiforov A.I. Identification of the permeability field for three-dimensional reservoir using the results of geophysical well survey. Georesursy, 2021, vol. 23 (1), pp. 106–111. (In Russ.) DOI: https://doi.org/10.18599/grs.2021.1.11
18. Soromotin A.V., Martyushev D.A., Melekhin A.A. Prediction of hydrodynamic parameters of the state of the bottomhole zone of wells using machine learning methods. Georesursy, 2024, vol. 26 (1), pp. 109–117. (In Russ.) DOI: https://doi.org/10.18599/grs.2024.1.9
19. Bakhmutov A.V., Kremenetskiy M.I., Nikonorova A.N., Ruzanova A.A. Diagnostics of contrast layered heterogeneity of productive reservoirs based on the results of pressure transient analysis. Georesursy, 2026. (In Russ.) DOI: https://doi.org/10.18599/grs.2026.4.7
20. Anh N.D. Rate-decline analysis for fracture-dominated shale reservoirs. SPE Canada Unconventional Resources Conference, 2014, vol. 14 (3), pp. 377–387. DOI: https://doi.org/10.2118/137748-PA
21. Chen Z. A shale gas resource potential assessment of Devonian Horn River strata using a well-performance method. Canadian Journal of Earth Sciences, 2016, vol. 53, no. 2, pp. 156–167. DOI: https://doi.org/10.1139/cjes-2015-0094
22. Chernyshov S.E., Popov S.N., Wang X., Derendyaev V.V., Yang Y., Liu H. Analysis of changes in the stress-strain state and permeability of a terrigenous reservoir based on a numerical model of the near-well zone with casing and perforation channels. Georesursy, 2024, vol. 26 (4), pp. 209–217. (In Russ.) DOI: https://doi.org/10.18599/grs.2024.4.6
23. Lawal A., Yang Y., He H., Baisa N.L. Machine learning in oil and gas exploration: a review. Ieee Access, 2024, vol. 12, pp. 19035–19058. DOI: http://doi.org/10.1109/access.2023.3349216
24. Al-Mudhafar W.J. Geostatistical simulation of facies and petrophysical properties for hetero-geneity modeling in A tidal depositional environment: a case study from upper shale member in A southern Iraqi oil field. Unconventional Resources Technology Conference. Houston, Texas, USA, 26–28 July 2021. pp. 2627–2642.
25. Thanh H.V., Sugai Y. Integrated modelling framework for enhancement history matching in fluvial channel sandstone reservoirs. Upstream Oil and Gas Technology, 2021, vol. 6, pp. 100027.
26. Jo S., Jeong H., Min B., Park C., Ki, Y., Kwon S., Sun A. Efficient deeplearning-based history matching for fluvial channel reservoirs. Journal of Petroleum Science and Engineering, 2022, vol. 208, article 109247.
27. Mahdy A., Zakaria W., Helmi A., Helaly A.S., Mahmoud A.M. Machine learning approach for core permeability prediction from well logs in Sandstone Reservoir, Mediterranean Sea, Egypt. Journal of Applied Geophysics, 2024, vol. 220, pp. 105249.
28. Zakharov L.А., Martyushev D.А., Ponomareva I.N. Predicting dynamic formation pressure using artificial intelligence methods. Journal of Mining Institute, 2022, vol. 253, pp. 23–32. (In Russ.) DOI: https://doi.org/10.31897/PMI.2022.11
29. Ponomareva I.N., Martyushev D.A. Reproduction of reservoir pressure in oil field development: prospects and problems of using methods machine learning. Georesursy, 2026, vol. 28 (1), pp. 43–53. (In Russ.) DOI: https://doi.org/10.18599/grs.2025.3.13
30. Davoodi S., Al-Shargabi M. Artificial intelligence techniques in oil and gas production technology: a review of applications, opportunities, and challenges. Petroleum Research, 2026. DOI: https://doi.org/10.1016/j.ptlrs.2026.08.003
31. Eremyan G., Magdeev A.M., Al-Shargabi M., Matveev I.V., Smirnov I.E., Shishaev G.Y., Davoodi S. Enhancing reservoir simulation history matching using SHAP value for parameter range selection. Energies, 2026, vol. 19, Article 2285. DOI: https://doi.org/10.3390/en19102285
32. Popov S., Chernyshov S., Gladkikh E. Experimental and numerical assessment of the influence of bottomhole pressure drawdown on terrigenous reservoir permeability and well productivity. Fluid Dynamics and Materials Processing, 2023, vol. 19 (3), pp. 619–634. DOI: https://doi.org/10.32604/fdmp.2022.021936
33. Martyushev D.A., Soromotin A.V., Pérez D.M., Rodriges A.G. A novel investigation of hydrodynamic data for permeability and skin factor estimation using machine learning models. Modeling Earth Systems and Environment, 2026, vol. 12 (5), Article 274. DOI: https://doi.org/10.1007/s40808-026-02933-2
34. Campos D., Wayo D.D.K., De Santis R.B., Martyushev D.A., Yaseen Z.M., Duru U.I., Saporetti C.M., Goliatt L. Evolutionary automated radial basis function neural network for multiphase flowing bottom-hole pressure prediction. Fuel, 2024, vol. 377, Article 132666. DOI: https://doi.org/10.1016/j.fuel.2024.132666
35. Ponomareva I.N., Galkin V.I., Martyushev D.A. Operational method for determining bottom hole pressure in mechanized oil producing wells, based on the application of multivariate regression analysis. Petroleum Research, 2021, vol. 6 (4), pp. 351–360. DOI: https://doi.org/10.1016/j.ptlrs.2021.05.010
36. Kumar I., Tripathi B.K., Singh A. Attention-based LSTM network-assisted time series forecasting models for petroleum production. Engineering Applications of Artificial Intelligence, 2023, vol. 123, Article 106440. DOI: https://doi.org/10.1016/j.engappai.2023.106440
