Operational monitoring of the well stock is a time-consuming task in base oil production. Among such tasks are the following: on-time detection of deviations in the flow rate of produced products, determination of the exact causes of changes in field parameters, as well as analysis of large data sets. This is especially true for oil and gas companies in case of large well stock and and high degree of reserves development. The development of digital technologies creates conditions for the integration of intelligent systems into the process of oil and gas production. Automatic processing of huge arrays of field data allows quickly monitoring the state of the base stock of wells for localization of the most problematic areas.
The article presents the developed new concept for working with oil flow rate deviations based on automated algorithms and mathematical approaches. The main objectives of the presented work are the development of a systematic approach to online monitoring field parameters, increasing accuracy and speed in determining deviations in oil production, as well as minimizing routine manual labor and reducing the impact of human factors. Due to the automated approach, the costs of developing the fund are significantly reduced, labor costs are reduced and the degree of influence of errors related to the human factor is reduced, in addition, daily monitoring helps to react in time to changes in the oil flow rate, and increasing the accuracy of explaining the causes of deviations significantly improves the quality of the selection of measures to eliminate them. According to methodology, software was implemented - a system of digital tools for identifying and explaining the causes of deviations in oil production, and factor analysis for the selected period. The developed approaches have been introduced into the process of the production company. This made it possible to increase the effectiveness of monitoring well stock.
References
1. Andrianova A.M., Yudin E.V., Ganeev T.A. et al., Application of intelligent methods for analysis high-frequency production data for solving oil engineering challenges (In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2021, no. 9, pp. 70–75, DOI: https://doi.org/10.24887/0028-2448-2021-9-70-75
2. Asmandiyarov R.N., Kladov A.E., Lubnin A.A. et al., Automatic approach to field data analysis (In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2011, no. 6, pp. 58–61.
3. Boschetti A., Massaron L., Python data science essentials, Packt Publishing Ltd., 2015, 472 p.
4. Mining J., Python machine learning: Everything you should know about Python Machine Learning including Scikit Learn, Numpy, PyTorch, Keras and Tensorflow with step-by-step examples, practical exercises, International Kindle Paperwhite, 2019, 124 p.
5. Pylov P.A., Ivina O.A., Fundamental'nye tipy kross-validatsii dlya otsenki kachestva modeley mashinnogo i glubokogo obucheniya (Fundamental types of cross-validation for evaluating the quality of machine learning and deep learning models), Proceedings of XIII All-Russian Scientific and Practical Conference of Young Scientists “ROSSIYa MOLODAYa” (RUSSIA YOUNG), 20-23 April 2021, URL: https://science.kuzstu.ru/wp-content/Events/Conference/RM/2021/RM21/pages/Articles/031520.pdf
