Applying Artificial Neural Network and XGBoost to Improve Data Analytics in Oil and Gas Industry
The application of machine learning and artificial intelligence is popular nowadays to improve data analytics in the oil and gas industry. A huge amount of data can be processed to gain insights about the subsurface conditions, even reducing time for manual review or interpretation. There are three cases to be discussed in this study that starts from porosity estimation of thin core image using Otsu's thresholding, estimation of oil production rate from sucker-rod pumping wells and sonic travel-time log generation. Two supervised learning algorithms are applied, XGBoost and Keras. These algorithms will capture all possible correlations between the input and output data. From data normalization, exploratory data analysis and model building, the workflow is built on Google Colab. The original dataset is split into training and testing. Tuning hyperparameters such as the number of hidden layers, neurons, activation function, optimizers and learning rates are captured to reduce the complexity of the model. The model is evaluated by error values and the coefficient of determination to estimate the model skill on unseen data.
Achmad, Z. (2017). Value creation of digital oilfield technology, study case Duri Field well monitoring. SPE/IATMI Asia Pacific Oil & Gas Conference and Exhibition. https://doi.org/10.2118/186903-ms
Bukar, I., Adamu, M. B., & Hassan, U. (2019). A machine learning approach to shear sonic log prediction. SPE Nigeria Annual International Conference and Exhibition. https://doi.org/10.2118/198764-ms
Campos, S., Baliño, J., Slobodcicov, I., Filho, D., & Paz, E. (2014). Orifice plate meter field performance: Formulation and validation in multiphase flow conditions. Experimental Thermal and Fluid Science, 58, 93-104. https://doi.org/10.1016/j.expthermflusci.2014.06.018
Caruana, R., & Niculescu-Mizil, A. (2006). An empirical comparison of supervised learning algorithms. Proceedings of the 23rd International Conference on Machine learning - ICML '06. https://doi.org/10.1145/1143844.1143865
Chen, T., & Guestrin, C. (2016). XGBoost. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. https://doi.org/10.1145/2939672.2939785
Doma, V., & Pirouz, M. (2020). A comparative analysis of machine learning methods for emotion recognition using EEG and peripheral physiological signals. Journal of Big Data, 7(1). https://doi.org/10.1186/s40537-020-00289-7
Khan, M. R., Alnuaim, S., Tariq, Z., & Abdulraheem, A. (2019). Machine learning application for oil rate prediction in artificial gas lift wells. SPE Middle East Oil and Gas Show and Conference. https://doi.org/10.2118/194713-ms
Li, J., Chow, T., & Ying-Lin Yu. (1995). The estimation theory and optimization algorithm for the number of hidden units in the higher-order feedforward neural network. Proceedings of ICNN'95 - International Conference on Neural Networks. https://doi.org/10.1109/icnn.1995.487330
Rygel, M. C. (2010). Photomicrograph of a quartz arenite (sandstone) from the St. Peter Formation (Ordovician) near Dixon, Illinois. https://commons.wikimedia.org/wiki/File:Quartz-arenite.jpg
Takacs, G. (2015). Sucker-rod pumping handbook: Production engineering fundamentals and long-stroke rod pumping. Gulf Professional Publishing.