Современные методы применения машинного обучения как инструмента прогнозирования добычи нефти
Сведения об авторах
- Университет Хазар, Баку, Азербайджанская Республика
- Университет Хазар, Баку, Азербайджанская Республика
- Санкт-Петербургский горный университет, Санкт-Петербург, Российская Федерация
- Азербайджанский университет архитектуры и строительства, Баку, Азербайджанская Республика
Аннотация
Прогнозирование добычи нефти играет важную роль в эффективной разработке месторождения нефти. Это помогает скорректировать действующую систему разработки месторождения. Детальное и точное прогнозирование уровня добычи нефти необходимо для оценки экономической и технологической эффективности разработки месторождения нефти. Прогнозирование уровня добычи можно осуществить различными способами. Одним из таких может быть использование специального программного обеспечения (tNavigator и др.). Использование данного программного обеспечения иногда сопряжено с длительными расчетами, поэтому для оперативного прогнозирования уровня добычи возможно использование других инструментов, таких как машинное обучение. Использование машинного обучения и искусственного интеллекта в нефтегазовой отрасли приобретает все большую популярность в последние годы, поскольку, используя исторические данные по добыче, возможно прогнозирование уровней добычи нефти/жидкости. Кроме того, для аналогичных целей могут быть использованы аналогичные месторождения со схожими геологическими характеристиками и историей эксплуатации. Помимо использования машинного обучения и искусственного интеллекта, в качестве инструмента прогнозирования возможно применение анализ кривой падения. Учитывая важность прогнозирования с точки зрения стратегического планирования, предлагается широкий спектр методов для получения точных прогнозов, основанных на характере доступных данных и вычислительной мощности. В данной статье представлен всесторонний анализ инструментов, используемых для долгосрочного прогнозирования добычи нефти, включая алгоритмы машинного обучения и анализ кривой падения добычи (DCA). Представлены результаты применения модели с долговременной и кратковременной памятью и ее практическая применимость на примере ее использования на скважине кандидате.
Ключевые слова
прогнозирование добычимашинное обучениекривая падения давленияискусственный интеллектсеть с долговременной и кратковременной памятью
Modern Methods of Using Machine Learning as a Tool for Oil Production Forecasting
Oil production forecasting plays an important role in efficient oil field development. This helps to adjust the current field development system. Detailed and accurate forecasting of oil production levels is necessary to assess the economic and technological efficiency of oil field development. Forecasting production levels can be done in various ways. One of these may be the use of special software (tNavigator, etc.). The use of this software sometimes involves lengthy calculations, so to quickly predict production levels, it is possible to use other tools, such as machine learning. The use of machine learning and artificial intelligence in the oil and gas industry has become increasingly popular in recent years, as by using historical production data, it is possible to predict oil/liquid production levels. In addition, similar deposits with similar geological characteristics and exploitation history can be used for similar purposes. In addition to using machine learning and artificial intelligence as a forecasting tool, it is possible to use decline curve analysis. Given the importance of forecasting from a strategic planning perspective, a wide range of methods have been proposed to obtain accurate forecasts based on the nature of available data and computing power. This article provides a comprehensive analysis of the tools used for long-term oil production forecasting, including machine learning algorithms and decline curve analysis (DCA), in particular. This article presents the results of applying the long- and short-term memory model and its practical applicability using the example of its use on a candidate well.
Keywords
production forecastingmachine learningpressure drop curveartificial intelligencelong short-term memory
Литература
- 1.Petrakov D.G. et al. Standards for Selection of Surfactant Compositions used in Completion and Stimulation Fluids. International Journal of Engineering, 2023, vol. 36, no. 9, pp. 1605-1610. DOI: 10.5829/JIE.2023.36.09C.03 Scholar
- 2.Aghaei H. et al. Density-dependent relationship between changes in ultrasonic wave velocities, effective stress, and petrophysical-elastic properties of sandstone. Ultrasonics, 2023, vol. 132, 106985 p. DOI: 10.1016/j.ultras.2023.106985 Scholar
- 3.Petrakov D.G., Penkov G.M., Zolotukhin A.B. Experimental study on the effect of rock pressure on sandstone permeability. Zapiski Gornogo institute, 2022, vol. 254, pp. 244-251. DOI: 10.31897/PMI.2022.24 Scholar
- 4.Negash B.M., Yaw A.D. Artificial neural network based production forecasting for a hydrocarbon reservoir under water injection. Petroleum Exploration and Development, 2020, T. 47, no. 2, pp. 383-392. DOI: 10.1016/S1876-3804(20)60055-5 Scholar
- 5.Bernert S.M., Shahzadani M. An imperialist competitive algorithm artificial neural network method to predict oil flow rate of the wells. International journal of computer applications, 2011, vol. 26, no. 10, pp. 47-50. DOI: 10.5120/3137-4326 Scholar
- 6.Falcone G., Alimonti C. The challenges of multiphase flow metering: today and beyond. International Conference on Offshore Mechanics and Arctic Engineering, 2007, vol. 42681, pp. 823-834. DOI: 10.1115/OMAE2007-29527 Scholar
- 7.Mirzaei-Paiaman A., Salavati S. The application of artificial neural networks for the prediction of oil production flow rate. Energy Sources, Part A: Recovery, utilization, and environmental effects, 2012, vol. 34, no. 19, pp. 1834-1843. DOI: 10.1080/15567036.2010.492386 Scholar
- 8.Gilbert W.E. Flowing and Gas-Lift Well Performance. Drilling and Production Practice, 1954, vol. 13, pp. 126-157. Scholar
- 9.Baxendell P.B. Bean performance-lake wells. Shell Internal Report. Houston, TX: Shell Oil, 1957. Scholar
- 10.Ros N.C.J. An analysis of critical simultaneous gas/liquid flow through a restriction and its application to flowmetering. Applied Scientific Research, 1960, vol. 9, pp. 374-388. DOI: 10.1007/BF00382215 Scholar
- 11.Achong L.B. Revised bean and performance formula for Lake Maracaibo wells. Shell Internal Report. Houston, TX: Shell Oil Co, 1961. Scholar
- 12.Mirzaei-Paiaman A., Salavati S. A new empirical correlation for sonic simultaneous flow of oil and gas through wellhead chokes for Persian oil fields. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 2013, vol. 35, no. 9, pp. 817-825. DOI: 10.1080/15567031003773304 Scholar
- 13.Zhang P.Y., Meng-Meng Z. The Application of BP Neural Network In Oil-Field. TELKOMNIKA Indonesian Journal of Electrical Engineering, 2013, vol. 11, no. 9, pp. 527-75283. DOI: 10.11591/telkomnika.v11i9.3280 Scholar
- 14.Liu W., Liu W.D., Gu J. Forecasting oil production using ensemble empirical model decomposition based Long Short-Term Memory neural network. Journal of Petroleum Science and Engineering, 2020, vol. 189, 107013 p. DOI: 10.1016/j.petrol.2020.107013 Scholar
- 15.Ali P.J.M. et al. Data normalization and standardization: a technical report. Mach Learn Tech Rep, 2014, vol. 1, no. 1, pp. 1-6. DOI: 10.13140/RG.2.2.28948.04489 Scholar
- 16.Rumelhart D.E., Hinton G.E., Williams R.J. Learning representations by back-propagating errors. Nature, 1986, vol. 323, no. 6088, pp. 533-536. DOI: 10.1038/323533a0 Scholar
- 17.Werbos P.J. Generalization of backpropagation with application to a recurrent gas market model. Neural networks, 1988, vol. 1, no. 4, pp. 339-356. DOI: 10.1016/0893-6080(88)90007-X Scholar
- 18.Elman J.L. Finding structure in time. Cognitive science, 1990, vol. 14, no. 2, pp. 179-211. DOI: 10.1016/0364-0213(90)90002-E Scholar
- 19.Bengio Y., Simard P., Frasconi P. Learning long-term dependencies with gradient descent is difficult. IEEE transactions on neural networks, 1994, vol. 5, no. 2, pp. 157-166. DOI: 10.1109/72.279181 Scholar
- 20.Hochreiter S., Schmidhuber J. Long short-term memory. Neural computation, 1997, vol. 9, no. 8, pp. 1735-1780. DOI: 10.1162/neco.1997.9.8.1735 Scholar
- 21.Gers F.A., Schmidhuber J., Cummins F. Learning to forget: Continual prediction with LSTM. Neural computation, 2000, vol. 12, no. 10, pp. 2451-2471. DOI: 10.1162/089976600300015015 Scholar
- 22.Staudemeyer R.C., Morris E.R. Understanding LSTM-a tutorial into long short-term memory recurrent neural networks. arXiv preprint arXiv:1909.09586, 2019. DOI: 10.48550/arXiv.1909.09586 Scholar
- 23.Understanding LSTM networks by Colah accessed at Understanding LSTM Networks -- colah's blog. Scholar
- 24.Werbos P.J. Backpropagation through time: what it does and how to do it, Proceedings of the IEEE, 1990, vol. 78, no. 10, pp. 1550-1560. DOI: 10.1109/5.58337 Scholar
- 25.Kohavi R. et al. A study of cross-validation and bootstrap for accuracy estimation and model selection. Ijcai, 1995, vol. 14, no. 2, pp. 1137-1145. Scholar
- 26.Berrar D. et al. Cross-Validation, 2019. DOI: 10.1016/B978-0-12-809633-8.20349-X Scholar
- 27.Kee E. et al. A Comparative Analysis of Cross-Validation Techniques for a Smart and Lean Pick-and-Place Solution with Deep Learning. Electronics, 2023, vol. 12, no. 11, 2371 p. DOI: 10.3390/electronics12112371 Scholar
- 28.Little M.A. et al. Using and understanding cross-validation strategies. Perspectives on Safe & Life Sciences, 2017, vol. 6, no. 5, gix020 p. DOI: 10.1093/gigascience/gix020 Scholar
- 29.Falessi D. et al. On the need of preserving order of data when validating within-project defect classifiers. Empirical Software Engineering, 2020, vol. 25, pp. 4805-4830. DOI: 10.1007/s10664-020-09868-x Scholar
- 30.Suradhaniwar S. et al. Time series forecasting of univariate agrometeorological data: a comparative performance evaluation via one-step and multi-step ahead forecasting strategies. Sensors, 2021, vol. 21, no. 7, pp. 2430. DOI: 10.3390/s21072430 Scholar
- 31.Zhang X., Zhang X. Optimal model averaging based on forward-validation. Journal of Econometrics, 2022. DOI: 10.1016/j.jeconom.2022.03.010 Scholar
- 32.Radzi S.F.M. et al. Hyperparameter tuning and pipeline optimization via grid search method and tree-based autoML in breast cancer prediction. Journal of personalized medicine, 2021, vol. 11, no. 10, 978 p. DOI: 10.3390/jpm11100978 Scholar
- 33.Belete D.M., Huchaih M.D. Grid search in hyperparameter optimization of machine learning models for prediction of HIV/AIDS test results. International Journal of Computers and Applications, 2022, vol. 44, no. 9, pp. 875-886. DOI: 10.1080/1206212X.2021.1974663 Scholar
- 34.Shekar B.H., Dagnew G. Grid search-based hyperparameter tuning and classification of microarray cancer data. 2019 second international conference on advanced computational and communication paradigms (ICACCP), IEEE, 2019, pp. 1-8. DOI: 10.1109/ICACCP.2019.8882943 Scholar
- 35.Bergstra J., Bengio Y. Random search for hyper-parameter optimization. Journal of machine learning research, 2012, vol. 13, no. 2. DOI: 10.5555/2188385.2188395 Scholar
- 36.Liashchynskyi P., Liashchynskyi P. Grid search, random search, genetic algorithm: a big comparison for NAS. arXiv preprint arXiv:1912.06059, 2019 DOI: 10.48550/arXiv.1912.06059 Scholar
- 37.St-Aubin P., Agard B. Precision and Reliability of Forecasts Performance Metrics. Forecasting, 2022, vol. 4, no. 4, pp. 882-903. DOI: 10.3390/forecast4040048 Scholar
- 38.Cerqueira V., Torgo L., Mozetič I. Evaluating time series forecasting models: An empirical study on performance estimation methods. Machine Learning, 2020, vol. 109, pp. 1997-2028. DOI: 10.1007/s10994-020-05910-7 Scholar
- 39.Chai T., Draxler R.R. Root mean square error (RMSE) or mean absolute error (MAE)? - Arguments against avoiding RMSE in the literature. Geoscientific model development, 2014, vol. 7, no. 3, pp. 1247-1250. DOI: 10.5194/gmd-7-1247-2014 Scholar
- 40.Park Y.H. Concise logarithmic loss function for robust training of anomaly detection model. arXiv preprint arXiv:2201.05748, 2022. DOI: 10.48550/arXiv.2201.05748 Scholar
- 1.Petrakov D.G. et al. Standards for Selection of Surfactant Compositions used in Completion and Stimulation Fluids. International Journal of Engineering, 2023, vol. 36, no. 9, pp. 1605-1610. DOI: 10.5829/JIE.2023.36.09C.03 Scholar
- 2.Aghaei H. et al. Density-dependent relationship between changes in ultrasonic wave velocities, effective stress, and petrophysical-elastic properties of sandstone. Ultrasonics, 2023, vol. 132, 106985 p. DOI: 10.1016/j.ultras.2023.106985 Scholar
- 3.Petrakov D.G., Penkov G.M., Zolotukhin A.B. Experimental study on the effect of rock pressure on sandstone permeability. Zapiski Gornogo institute, 2022, vol. 254, pp. 244-251. DOI: 10.31897/PMI.2022.24 Scholar
- 4.Negash B.M., Yaw A.D. Artificial neural network based production forecasting for a hydrocarbon reservoir under water injection. Petroleum Exploration and Development, 2020, T. 47, no. 2, pp. 383-392. DOI: 10.1016/S1876-3804(20)60055-5 Scholar
- 5.Berneti S.M., Shahbazian M. An imperialist competitive algorithm artificial neural network method to predict oil flow rate of the wells. International journal of computer applications, 2011, vol. 26, no. 10, pp. 47-50. DOI: 10.5120/3137-4326 Scholar
- 6.Falcone G., Alimonti C. The challenges of multiphase flow metering: today and beyond. International Conference on Offshore Mechanics and Arctic Engineering, 2007, vol. 42681, pp. 823-834. DOI: 10.1115/OMAE2007-29527 Scholar
- 7.Mirzaei-Paiaman A., Salvavati S. The application of artificial neural networks for the prediction of oil production flow rate. Energy Sources, Part A: Recovery, utilization, and environmental effects, 2012, vol. 34, no. 19, pp. 1834-1843. DOI: 10.1080/15567036.2010.492386 Scholar
- 8.Gilbert W.E. Flowing and Gas-Lift Well Performance. Drilling and Production Practice, 1954, vol. 13, pp. 126-157. Scholar
- 9.Baxendell P.B. Bean performance-lake wells. Shell Internal Report. Houston, TX: Shell Oil, 1957. Scholar
- 10.Ros N.C.J. An analysis of critical simultaneous gas/liquid flow through a restriction and its application to flowmetering. Applied Scientific Research, 1960, vol. 9, pp. 374-388. DOI: 10.1007/BF00382215 Scholar
- 11.Achong L.B. Revised bean and performance formula for Lake Maracaibo wells. Shell Internal Report. Houston, TX: Shell Oil Co, 1961. Scholar
- 12.Mirzaei-Paiaman A., Salvavati S. A new empirical correlation for sonic simultaneous flow of oil and gas through wellhead chokes for Persian oil fields. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 2013, vol. 35, no. 9, pp. 817-825. DOI: 10.1080/15567031003773304 Scholar
- 13.Zhang P.Y., Meng-Meng Z. The Application of BP Neural Network In Oil-Field. TELKOMNIKA Indonesian Journal of Electrical Engineering, 2013, vol. 11, no. 9, pp. 527-75283. DOI: 10.11591/telkomnika.v11i9.3280 Scholar
- 14.Liu W., Liu W.D., Gu J. Forecasting oil production using ensemble empirical model decomposition based Long Short-Term Memory neural network. Journal of Petroleum Science and Engineering, 2020, vol. 189, 107013 p. DOI: 10.1016/j.petrol.2020.107013 Scholar
- 15.Ali P.J.M. et al. Data normalization and standardization: a technical report. Mach Learn Tech rep, 2014, vol. 1, no. 1, pp. 1-6. DOI: 10.13140/RG.2.2.28948.04489 Scholar
- 16.Rumelhart D.E., Hinton G.E., Williams R.J. Learning representations by back-propagating errors. Nature, 1986, vol. 323, no. 6088, pp. 533-536. DOI: 10.1038/323533a0 Scholar
- 17.Werbos P.J. Generalization of backpropagation with application to a recurrent gsmarket model. Neural networks, 1988, vol. 1, no. 4, pp. 339-356. DOI: 10.1016/0893-6080(88)90007-X Scholar
- 18.Elman J.L. Finding structure in time. Cognitive science, 1990, vol. 14, no. 2, pp. 179-211. DOI: 10.1016/0364-0213(90)90002-E Scholar
- 19.Bengio Y., Simard P., Frasconi P. Learning long-term dependencies with gradient descent is difficult. IEEE transactions on neural networks, 1994, vol. 5, no. 2, pp. 157-166. DOI: 10.1109/72.279181 Scholar
- 20.Hochreiter S., Schmidhuber J. Long short-term memory. Neural computation, 1997, vol. 9, no. 8, pp. 1735-1780. DOI: 10.1162/neco.1997.9.8.1735 Scholar
- 21.Gers F.A., Schmidhuber J., Cummins F. Learning to forget: Continual prediction with LSTM. Neural computation, 2000, vol. 12, no. 10, pp. 2451-2471. DOI: 10.1162/089976600300015015 Scholar
- 22.Staudemeyer R.C., Morris E.R. Understanding LSTM–a tutorial into long short-term memory recurrent neural networks. arXiv preprint arXiv:1909.09586, 2019. DOI: 10.48550/arXiv.1909.09586 Scholar
- 23.Understanding LSTM networks by Colah accessed at Understanding LSTM Networks – colah's blog. Scholar
- 24.Werbos P.J. Backpropagation through time: what it does and how to do it. Proceedings of the IEEE, 1990, vol. 78, no. 10, pp. 1550-1560. DOI: 10.1109/5.58337 Scholar
- 25.Kohavi R. et al. A study of cross-validation and bootstrap for accuracy estimation and model selection. Jicai, 1995, vol. 14, no. 2, pp. 1137-1145. Scholar
- 26.Berrar D. et al. Cross-Validation, 2019. DOI: 10.1016/B978-0-12-809633-8.20349-X Scholar
- 27.Kee E. et al. A Comparative Analysis of Cross-Validation Techniques for a Smart and Lean Pick-and-Place Solution with Deep Learning. Electronics, 2023, vol. 12, no. 11, 2371 p. DOI: 10.3390/electronics12112371 Scholar
- 28.Little M.A. et al. Using and understanding cross-validation strategies. Perspectives on Saeb et al. GigaScience, 2017, vol. 6, no. 5, gix020 p. DOI: 10.1093/gigascience/gix020 Scholar
- 29.Falessi D. et al. On the need of preserving order of data when validating within-project defect classifiers. Empirical Software Engineering, 2020, vol. 25, pp. 4805-4830. DOI: 10.1007/s10664-020-09868-x Scholar
- 30.Suradhaniwar S. et al. Time series forecasting of univariate agrometeorological data: a comparative performance evaluation via one-step and multi-step ahead forecasting strategies. Sensors, 2021, vol. 21, no. 7, pp. 2430. DOI: 10.3390/s21072430 Scholar
- 31.Zhang X., Zhang X. Optimal model averaging based on forward-validation. Journal of Econometrics, 2022. DOI: 10.1016/j.jeconom.2022.03.010 Scholar
- 32.Radzi S.F.M. et al. Hyperparameter tuning and pipeline optimization via grid search method and tree-based autoML in breast cancer prediction. Journal of personalized medicine, 2021, vol. 11, no. 10, 978 p. DOI: 10.3390/jpm11100978 Scholar
- 33.Belete D.M., Huchaih M.D. Grid search in hyperparameter optimization of machine learning models for prediction of HIV/AIDS test results. International Journal of Computers and Applications, 2022, vol. 44, no. 9, pp. 875-886. DOI: 10.1080/1206212X.2021.1974663 Scholar
- 34.Shekar B.H., Dagnew G. Grid search-based hyperparameter tuning and classification of microarray cancer data. 2019 second international conference on advanced computational and communication paradigms (ICACCP), IEEE, 2019, pp. 1-8. DOI: 10.1109/ICACCP.2019.8882943 Scholar
- 35.Bergstra J., Bengio Y. Random search for hyper-parameter optimization. Journal of machine learning research, 2012, vol. 13, no. 2. DOI: 10.5555/2188385.2188395 Scholar
- 36.Liashchynskyi P., Liashchynskyi P. Grid search, random search, genetic algorithm: a big comparison for NAS. arXiv preprint arXiv:1912.06059, 2019. DOI: 10.48550/arXiv.1912.06059 Scholar
- 37.St-Aubin P., Agard B. Precision and Reliability of Forecasts Performance Metrics. Forecasting, 2022, vol. 4, no. 4, pp. 882-903. DOI: 10.3390/forecast4040048 Scholar
- 38.Cerqueira V., Torgo L., Mozetič I. Evaluating time series forecasting models: An empirical study on performance estimation methods. Machine Learning, 2020, vol. 109, pp. 1997-2028. DOI: 10.1007/s10994-020-05910-7 Scholar
- 39.Chai T., Draxler R.R. Root mean square error (RMSE) or mean absolute error (MAE)? - Arguments against avoiding RMSE in the literature. Geoscientific model development, 2014, vol. 7, no. 3, pp. 1247-1250. DOI: 10.5194/gmd-7-1247-2014 Scholar
- 40.Park Y.H. Concise logarithmic loss function for robust training of anomaly detection model. arXiv preprint arXiv:2201.05748, 2022. DOI: 10.48550/arXiv.2201.05748 Scholar
Другие статьи выпуска
Весь выпуск (Том 24, № 1, 2024)- Оценка влияния тектонических факторов на формирование глубокопогруженных нефтегазоматеринских пород Верхнепечорской впадиныс. 2–9
- К вопросу о дифференциации рассеянного органического вещества Верхнедевонско-Турнейской толщи территории Пермского краяс. 10–17
- Анализ и снижение неопределенности параметра проницаемости при моделировании карбонатного резервуарас. 18–26
- Определение свойств глинистых пород-коллекторов на различных стадиях разработки месторождений нефтис. 27–34
- Численное моделирование поверхностно-активных веществ полимерного заводнения совместно с заводнением слабосоленой водой в Matlab на примере месторождения «Нефть Дашлары»с. 35–43
Информация о статье
- Опубликована
- 31.03.2024
- В выпуске
- 31.03.2024
- Выпуск
- Том 24, № 1 (2024)