|
The paper proposes a methodology for applying machine-learning methods under selecting the candidate wells to hydraulic fracturing one of the Rosneft Oil Company. Currently, a large amount of information is being collected during field development, the analysis of which by traditional methods is practically impossible due to the large time labor involved in processing and making decisions. In recent years, such problems have been solved using modern digital technologies. The most actively developing methods are data analysis based on machine learning algorithms intended to extract knowledge from the presented data array in order to make decisions regarding the objects under consideration. The goal of the paper was to develop an integrated approach for the selection of candidate wells for hydraulic fracturing. As part of this goal, predictive machine learning models were created for the following indicators: starting oil and liquid production rates, oil production rates after 1, 3, 6 months, and an economic profitability indicator. The 3 machine learning algorithms were used to forecast each indicator. The algorithm showed the smallest error was chosen for each model. In the course of the work, it was shown that for modeling target indicators after the hydraulic fracturing nonlinear gradient boosting and random forest algorithms turned out to be the best. Selection of candidate wells for hydraulic fracturing operations is carried out based on a ranked list of candidate wells according to the forecasted target technological and economic indicators. Testing the proposed approach at one of the Rosneft's fields of the developed approach has shown the potential to improve forecasting accuracy and economic efficiency, which will potentially make it possible to increase the efficiency of the Rosneft's field development in the future. References 1. Aryanto A., Kasmungin S., Fathaddin M.T., Hydraulic fracturing candidate-well selection using artificial intelligence approach, Prosiding Seminar Nasional Cendekiawan (Proceedings of the National Scholar Seminar, 2018, pp. 1–7. 2. Yanfang W., Salehi S., Refracture candidate selection using hybrid simulation with neural network and data analysis techniques, Journal of Petroleum Science and Engineering, 2014, V. 123, pp. 138–146. 3. Rahmanifard H., Plaksina T., Application of artifcial intelligence techniques in the petroleum industry: a review, Artifcial Intelligence Review, 2018, pp. 1–24. 4. Mohaghegh S., Reeves S., Hill D. et al., Development of an intelligent systems approach for restimulation candidate selection, SPE-59767-MS, 2000. 5. Ting Yu et al., Comparison of candidate-well selection mathematical models for hydraulic fracturing, Fuzzy Systems & Operations Research and Management, 2015, V. 367, 289 p. 6. Galiullin M.M., Shabarov A.B., Application of the fuzzy sets theory for selection of wells with a view to wellwork on oil fields (In Russ.), Vestnik Tyumenskogo gosudarstvennogo universiteta = Tyumen State University Herald, 2011, no. 7, pp. 30–37. 7. Alimkhanov R., Samoylova I., Application of data mining tools for analysis and prediction of hydraulic fracturing efficiency for the BV8 reservoir of the Povkh oil field (In Russ.), SPE-171332-MS, 2014, DOI:10.2118/171332-MS. 8. Davletova A.R., Kolonskikh A.V., Fedorov A.I., Fracture reorientation of secondary hydraulic fracturing operation (In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2017, no. 11, pp. 110–113. 9. Savchenko P.D., Fedorov A.I., Kolonskikh A.V. et al., Method for selecting well candidates based on the effect of fracture reorientation (In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2017, no. 11, pp. 114–117. 10. Zdol'nik S.E., Nekipelov Yu.V., Gaponov M.A., Folomeev A.E., Introduction of innovative hydrofracturing technologies on carbonate reservoirs of Bashneft PJSOC (In Russ.), Neftyanoe khozyaystvo = Oil Industry, 2016, no. 7, pp. 92–95. |
