Review of neural network models for solving the problems of predicting emergency situations and ensuring the safe operation of oil and gas wells

An analytical study was carried out on the problem of preventing emergency situations and predictive diagnostics of equipment during hydrocarbon production in oil and gas fields as well as the ways to solve this problem by means of artificial intelligence based on deep neural networks. One of the key factors hindering the development of predictive equipment diagnostic systems is the lack of data describing pre-emergency situations, which is necessary for high-quality training of neural network models. An analysis of recent publications and research on the subject of telemetry data analysis and emergency recognition is provided. Neural network models are considered that can be used to predict the failure of pumping and compressor equipment and other units. Cases of the use of neural network models specially trained to solve this problem, as well as neural network models used in other tasks but analyzing similar data structures, were studied. The issue of transfer learning is raised to adapt neural network models originally developed and trained for other areas to use in the area under consideration in order to reduce the sample size when training industrial artificial intelligence. A comparison of the achieved results was carried out, and the advantages and disadvantages of existing technical solutions were identified.

1. Geeta Y., Kolin P. Architecture and security of SCADA systems. International Journal of Critical Infrastructure Protection. 2021;34:100433. DOI: 10.1016/j.ijcip.2021.100433.

2. Tcharo H., Vorobiev А.E. Digitalization of the oil industry: basic approaches and justification of “smart” technologies. Vestnik evraziyskoy nauki. 2018;10(2):77. (In Russ.).

3. Anirbid S., Kriti Y. Application of machine learning and artificial intelligence in oil and gas industry. Petroleum Research. 2021;6(4):379–391. DOI: 10.1016/j.ptlrs.2021.05.009.

4. Hamidulin T.G. Primenenie iskustvennih neyronnykh setey. Ekonomika i socium. 2017;38(7):313–318. (In Russ.).

5. Zhou D., Huang D. Vibration-based fault diagnosis of the natural gas compressor using adaptive stochastic resonance realized by Generative Adversarial Networks. Engineering Failure Analysis. 2020;116:104759. DOI: 10.1016/j.engfailanal.2020.104759.

6. Topol'nikov A.S Mashinnoe obuchenie dlya mekhanizirovannoy dobychi nefti. Delovoy zhurnal Neftegaz.RU. 2021;5:14–19. URL: https://magazine.neftegaz.ru/articles/dobycha/682013-mashinnoe-obuchenie-dlya-mekhanizirovannoy-dobychi-nefti [accessed on 27.10.2023].

7. Wong P., Wong W.K. A minimalist approach for detecting sensor abnormality in oil and gas platforms. Petroleum Research. 2022;7(2):177–185. DOI: 10.1016/j.ptlrs.2021.09.007.

8. Vershinin V.E. Neyrosetevoe modelirovanie: prognozirovanie pokazateley dobychi skvazhin v usloviyakh nestatsionarnogo zavodneniya. Delovoy zhurnal Neftegaz.RU. 2022;5:26–32. URL: https://magazine.neftegaz.ru/articles/tsifrovizatsiya/740217-neyrosetevoe-modelirovanie-prognozirovanie-pokazateley-dobychi-skvazhin-v-usloviyakh-nestatsionarnog/?ysclid=ldwqmqk0bh84798351 [accessed on 27.10.2023]. (In Russ.).

9. Dyer A.S., Zaengle D. Applied machine learning model comparison: Predicting offshore platform integrity with gradient boosting algorithms and neural networks. Marine Structures. 2022;83:103152. DOI: 10.1016/j.marstruc.2021.103152.

10. Kozlenko M., Kuz M. Fault diagnosis of natural gas pumping unit based on machine learning. 6th International Scientific and Practical Conference on Applied Systems and Technologies in the Information Society.2022;4:271. DOI: 10.5281/zenodo.7409180.

11. Wu Y., Feng Z. Fault diagnosis algorithm of beam pumping unit based on transfer learning and DenseNet model. Applied sciences. 2022;21(12):11091. DOI: 10.3390/app122111091.

12. Yolchuyev A. Feed-forward neural network based petroleum wells equipment failure prediction. Engineering. 2023;15(3):163–175. DOI: 10.4236/eng.2023.153013.

13. Li Y., Ge T. Imminence monitoring of critical events: a representation learning approach. In: Proceedings of the 2021 International Conference on Management of Data (SIGMOD '21). Association for Computing Machinery, New York, USA, 2021. p. 1103–1115. DOI: 10.1145/3448016.3452804

14. Carvalho B.G. Evaluating machine learning techniques for detection of flow instability events in offshore oil wells. Universidade Federal do Espírito Santo. 2021;1:1–59.

15. Marins M.A., Barros B.D. Fault detection and classification in oil wells and production/service lines using random forest. Journal of Petroleum Science and Engineering. 2021;197:107879. DOI: 10.1016/j.petrol.2020.107879.

16. Asyaev G.D., Sokolov А.N. Predictive information protection models of automated water management system based on time series using machine learning technologies. Vestnik UrFO Bezopasnost' v informatsionnoy sfere = Vestnik UrFO. Security in the Information Sphere. 2021;42(4):39–45. (In Russ.).

17. Marushko E.E., Doudkin A.A. Ensembles of neural networks for forecasting of time series of spacecraft telemetry. Optical Memory and Neural Networks. 2017;26(1):47–54. DOI: 10.3103/S1060992X17010064.

18. Jain R., Rohit M. Prediction of telemetry data using machine learning techniques. International Journal of Engineering Research & Technology. 2022;11(9). DOI: 10.17577/IJERTV11IS090048.

19. Wu H., Xu J. Autoformer: decomposition transformers with auto-correlation for long-term series forecasting. Arxiv. 2022;5. DOI:0.48550/arXiv.2106.13008.

20. Zhang K., Wang S.C. Anomaly detection of control moment gyroscope based on working condition classification and transfer learning. Applied sciences. 2023;13(7):4259–9. DOI: 10.3390/app13074259.

21. Zhou T., Ziqing M. FEDformer: frequency enhanced decomposed transformer for long-term series forecasting. Arxiv. 2022;3. DOI: 10.48550/arXiv.2201.12740.

22. Kuang L., Pobbathi S. Predicting age and gender from network telemetry: Implications for privacy and impact on policy. PLoS ONE 2022;17(7):e0271714. DOI: 10.1371/journal.pone.0271714.

23. Wibawa A.P., Elmunsyah H. Time-series analysis with smoothed Convolutional Neural Network. Journal of Big Data. 2022;9(1). DOI: 10.1186/s40537-022-00599-y

24. Özmen Ö., Sinanoğlu C. Prediction of leakage from an axial piston pump slipper with circular dimples using deep neural networks. Chinese Journal of Mechanical Engineering. 2020;33(1). DOI: 10.1186/s10033-020-00443-5.

25. Yang L., Ma Y. Improved deep learning-based telemetry data anomaly detection to enhance spacecraft operation reliability. Microelectronics Reliability. 2021;126. DOI: 10.1016/j.microrel.2021.114311.

26. Ibrahim S., Ayman A. Machine learning techniques for satellite fault diagnosis. Ain Shams Engineering Journal. 2020;11(1). DOI: 10.1016/j.asej.2019.08.006

27. Skobtsov V., Sokolov B. Hybrid neural network models in the problem of telemetry data multiclass classification of small spacecrafts. Vestnik VGU. Seriya: Sistemnyi analiz i informatsionnye tekhnologii = Proceedings of Voronezh State University. Series: Systems Analysis and Information Technologies. 2022;(3):99. DOI: 10.17308/sait/1995-5499/2022/3/99-114. (In Russ.).