Prediction of incidents occurrence at injection wells using machine learning algorithms
UDC: 622.276
DOI: 10.33285/0207-2351-2023-9(657)-16-21
Authors:
MEKHONOSHIN ROMAN O.
1,
VILDANOV TIMUR F.1,
KORDIK KIRILL E.2,
CHERNOBROVIN EVGENY V.1,
YAMLIKHIN RADIK R.3,4,
ZINATULLIN ILGAM A.3,4,
ELIZAROV ALEXEY V.3
1 LUKOIL-Engineering, Tyumen, Russia
2 LUKOIL-Engineering Limited KogalymNIPIneft Branch Office in Tyumen, Tyumen, Russia
3 LUKOIL-Western Siberia, Kogalym, Russia
4 Pokachevneftegas
Keywords: machine learning, forecasting of emergency incidents at a well, inspection of tubing, training database, machine learning algorithm
Annotation:
The article presents the results of a pilot project implementation that envisages creation of a digital instrument based on machine learning algorithms in order to predict emergency incidents at injection wells. The approaches to a database formation for a machine learning model by differentiating and converting a number of parameters into integral values are described. The considered case is an example of the application of a machine learning model as a predictive analytics tool for optimizing the costs of an oil and gas producing enterprise associated with repair work on injection wells. The reliability of the forecast obtained using the created model was confirmed during a scheduled inspection of tubing in 2023 at a pilot site of one of the company’s territorial production enterprises.
Bibliography:
1. Mashinnoe obuchenie / Kh. Brink, Dzh. Richards, M. Feverolf [i dr.] – SPb.: Piter, 2017. – 336 s.
2. Myuller A., Gvido S. Vvedenie v mashinnoe obuchenie s pomoshch’yu Python. – M.: Vil’yams, 2017. – 480 s.
3. Advanced Machine Learning Methods for Production Data Pattern Recognition / N. Subrahmanya [et al.] // Intelligent Energy Conference and Exhibition held in Utrecht, the Netherlands – Society of Petroleum Engineers, 2014.
4. Accelerating field optimization for Shell in the Neuquén Basin using Novi Labs machine learning models and data analytics / S. McEntyre [et al.] // Unconventional Resources Technology Conference held in Houston, Texas, USA – Society of Petroleum Engineers, 2022.
5. Machine Learning Engine for Real-Time ESP Failure Detection and Diagnostics / R. Adolfo [et al.] // Middle East Artificial Lift Conference and Exhibition held in Manama, Bahrain – Society of Petroleum Engineers, 2022.
6. Primenenie metodov mashinnogo obucheniya dlya prognozirovaniya veroyatnosti ostanovok dobyvayushchikh skvazhin na osnove parametrov rezhimov ikh ekspluatatsii / S.A. Yarikov [i dr.] // Neftegazovyy inzhiniring. – 2022. – № 5. – S. 85–89.
7. Novyy podkhod k doutochneniyu prognozov proksi-modeley plasta s pomoshch’yu algoritmov mashinnogo obucheniya / O.V. Zotkin [i dr.] // Informatsionnye tekhnologii. – 2019. – № 12. – S. 60–63.
8. Boslaf S. Statistika dlya vsekh. – M.: DMK-Press, 2017. – 586 s.
Back to issue