No.4(44),2025 PP.266-283

Neftyanaya provintsiya No. 4(44), 2025

Predicting of gas wells waterсut using a machine learning models (ML-models)

A.Yu. Yushkov, N.P. Lychagin, R.Yu. Shumeiko, N.M. Ogurechnikov, V.S. Shumalkin
DOI: https://doi.org/10.25689/NP.2025.4.266-283

Abstract

The article discusses the problem of prediction accuracy of gas well watercut, its relevance, and existing solutions. As an alternative prediction method, the authors suggest using machine learning tools (ML model). The training case is based on synthetic data obtained from the results of hydrodynamic modeling of the development of gas deposits of various configurations with low gas saturated thickness. 13 previously known geological and technological factors characterizing each well are accepted as input parameters of the ML model. The output (forecast) parameters of the model are: the year of the beginning of well watercut and the growth dynamics of the well water-gas ratio. The effectiveness of various machine learning algorithms is analyzed, and the implementation of the XGBoost algorithm is considered in more detail. Based on the results of testing the model on a control sample, a good accuracy of forecasting the time of the beginning of well flooding was obtained. It is concluded that ML models are capable of solving specific development management tasks, providing a quick forecast of selected target parameters.

Key words:

gas field development, gas well, hydrodynamic modelling, watercut prediction, machine learning, neural networks

References

1. Ponomarev A.I., Shayahmetov A.I. Prognozirovanie obvodneniya fonda dobivayuschih skvajin na krupnih gazovih mestorojdeniyah. [Forecasting the Watering of Production Wells in Large Gas Fields]. Ufa [Ufa], 2014. – 144 p. (in Russian)

2. Usachev V.D. Razrabotka metodov prognoza obvodneniya gazovih zalejei i skvajin na baze kompozicionnih gidrodinamicheskih modelei [Development of Methods for Forecasting Watering of Gas Deposits and Wells Based on Composite Hydrodynamic Models]. Moskva: dis. kand. tehn. nauk [Moscow: dissertation of the Candidate of Technical Sciences], 2018. – 116 p. (in Russian)

3. Usachev V.D., Peremishcev Yu.A. Prognozirovanie obvodneniya gazovih zalejei vodoplavayuschego tipa na primere mestorojdenii Zapadnoi Sibiri [Forecasting of Watering of Gas Deposits of the Water-Swimming Type on the Example of Deposits in Western Siberia]. Neftepromislovoe delo [petroleum engineering], 2018. №1. P. 12-17. (in Russian).

4. Gasumov E.R. Prognozirovanie vremeni obvodneniya i samozadavlivaniya gazovih skvajin (na primere senomanskoi zaleji) [Prediction of the Time of Watering and Self-Pressure of Gas Wells (Using the Cenoman Deposit as an Example)]. Evraziiskii Soyuz Uchenih [Eurasian Union of Scientists], 2020. №8(77). С. 19-22. (in Russian)

5. Milovanova V.V., Ibragimova D.R., Zaharov R.V., Krasnoborov S.V. Prognozirovanie obvodneniya gazovih skvajin_ novii podhod s ispolzovaniem promislovih dannih [Predicting Water Cut in Gas Wells: A New Approach Using Field Data]. Injener_neftyanik [Petroleum Engineer], 2021. №1. P. 5-10. (in Russian)

6. Kislicin A.A., Kuznecov S.V., Podnebesnih A.V., Granovskii A.M. Ispolzovanie iskusstvennih neironnih setei dlya prognozirovaniya dinamiki obvodneniya gorizontalnih skvajin [Using artificial neural networks to predict the dynamics of water cut in horizontal wells]. Tyumen: Vestnik TyumGU. Fiziko_matematicheskoe modelirovanie. Neft_ gaz_ energetika [Tyumen: Tyumen State University Bulletin. Physical and Mathematical Modeling. Oil, Gas, and Energy], 2019. Volume 5. №4(20). P. 160-180. (in Russian)

7. Ibraev A.E._ Kamaridenova G.S._ Baluanov B.A._ Elemesov A.S. Prognoz obvodnennosti novih skvajin s pomoschyu mashinnogo obucheniya [Predicting the water cut of new wells using machine learning]. Vestnik neftegazovoi otrasli Kazahstana [Bulletin of the oil and gas industry of Kazakhstan], 2023. - Volume 5, №3. P.20-34. (in Russian)

8. Ibrahim N.M., Alharbi A.A., Alzahrani T.A., Abdulkarim A.M., Alessa I.A., Hameed A.M., Albabtain A.S., Alqahtani D.A., Alsawwaf M.K., Almuqhim A.A. Well Performance Classification and Prediction: Deep Learning and Machine Learning Long Term Regression Experiments on Oil, Gas, and Water Production // Sensors. 2022.

Authors

A.Yu. Yushkov, candidate of technical sciences, senior expert on gas field development, Tyumen Petroleum Research Center LLC, Tyumen Industrial University
79/1, Osipenko St., 625002, Tyumen, Russian Federation
E-mail: ayyushkov@tnnc.rosneft.ru

N.P. Lychagin, specialist student, Tyumen Industrial University
38, Volodarsky St., 625000, Tyumen, Russian Federation
E-mail: kolia_lichagin2003@mail.ru

R.Yu. Shumeiko, specialist student, Tyumen Industrial University
38, Volodarsky St., 625000, Tyumen, Russian Federation
E-mail: russhum@bk.ru

N.M. Ogurechnikov, bachelor's degree student, Tyumen Industrial University
38, Volodarsky St., 625000, Tyumen, Russian Federation
E-mail: nikitka.ogurechnikov@mail.ru

V.S. Shumalkin, bachelor's degree student, Tyumen Industrial University
38, Volodarsky St., 625000, Tyumen, Russian Federation
E-mail: shymalkin3000@gmail.com

For citation:

A.Yu. Yushkov, N.P. Lychagin, R.Yu. Shumeiko, N.M. Ogurechnikov, V.S. Shumalkin Prognozirovaniye obvodneniya gazovykh skvazhin pri pomoshchi modeley mashinnogo obucheniya (ML-modeley) [Predicting of gas wells waterсut using a machine learning models (ML-models)]. Neftyanaya Provintsiya, No. 4(44), 2025. pp. 266-283. DOI https://doi.org/10.25689/NP.2025.4.266-283. EDN TGVEVL (in Russian)

© Non-governmental organization Volga-Kama Regional Division of the Russian Academy of Natural Science, 2015-2025 All the materials of the journal are available under the Creative Commons Attribution 4.0 License (https://creativecommons.org/licenses/by/4.0/)

Predicting of gas wells waterсut using a machine learning models (ML-models) A.Yu. Yushkov, N.P. Lychagin, R.Yu. Shumeiko, N.M. Ogurechnikov, V.S. Shumalkin DOI: https://doi.org/10.25689/NP.2025.4.266-283

Key words: gas field development, gas well, hydrodynamic modelling, watercut prediction, machine learning, neural networks

Predicting of gas wells waterсut using a machine learning models (ML-models)

A.Yu. Yushkov, N.P. Lychagin, R.Yu. Shumeiko, N.M. Ogurechnikov, V.S. Shumalkin
DOI: https://doi.org/10.25689/NP.2025.4.266-283

Key words:

gas field development, gas well, hydrodynamic modelling, watercut prediction, machine learning, neural networks