Prediksi Tekanan Pori Berdasarkan Data Logging Sumur Menggunakan Deep Neural Network
Article Sidebar
Main Article Content
Abstract
Tekanan pori formasi merupakan data penting yang digunakan dalam mendesain parameter dalam operasi pengeboran. Kesalahan dalam menentukan tekanan pori formasi dapat menimbulkan permasalahan dalam pengeboran yang dapat menyebabkan bertambahnya biaya pengeboran hingga kehilangan nyawa pekerja. Tekanan pori bisa didapatkan melalui pengukuran langsung maupun menggunakan metode empiris. Pengukuran langsung tidak dilakukan di setiap kedalaman karena biayanya yang tinggi. Metode empiris memiliki keterbatasan dalam menentukan tekanan pori formasi dikarenakan metode ini memerlukan adanya keakuratan dalam analisis Normal Compaction Trendline dan batasan untuk digunakan pada formasi tertentu. Penelitian ini menggali potensi jaringan saraf tiruan untuk memprediksi tekanan pori berdasarkan data logging sumur. Dengan arsitektur jaringan saraf tiruan yang menggunakan tiga hidden layer, model dibangun dengan lima fungsi aktivasi dan metriks evaluasi Mean Absolute Error. Semua mampu memprediksi tekanan pori dengan baik yang ditunjukkan dengan nilai loss dibawah 0.5. Diantara kelima fungsi aktivasi, Exponential Linear Unit menghasilkan loss paling baik sebesar 0.07 dibanding model dengan fungsi aktivasi lainnya.
Downloads
Article Details
References
[2] M. D. Zoback, Reservoir Geomechanics. Cambridge, United Kingdom: CAMBRIDGE UNIVERSITY PRESS, 2007.
[3] M. E. C. and F. S. Y. J. A.T. Bourgoyne Jr, K.K. Millheim, Applied Drilling Engineering. Society of Petroleum Engineers, 1991.
[4] C. E. Hottmann and R. K. Johnson, “Estimation of Formation Pressures from Log-Derived Shale Properties,” J. Pet. Technol., vol. 17, no. 06, pp. 717–722, Jun. 1965.
[5] B. A. Eaton, “The Equation for Geopressure Prediction from Well Logs,” in Fall Meeting of the Society of Petroleum Engineers of AIME, 1975.
[6] G. L. Bowers, “Pore Pressure Estimation From Velocity Data: Accounting for Overpressure Mechanisms Besides Undercompaction,” SPE Drill. Complet., vol. 10, no. 02, pp. 89–95, Jun. 1995.
[7] J. Zhang, “Pore pressure prediction from well logs: Methods, modifications, and new approaches,” Earth-Science Rev., vol. 108, no. 1–2, pp. 50–63, Sep. 2011.
[8] R. E. Swarbrick, “Pore-Pressure Prediction: Pitfalls in Using Porosity,” in Offshore Technology Conference, 2001.
[9] A. Ahmed, S. Elkatatny, A. Ali, and A. Abdulraheem, “Comparative analysis of artificial intelligence techniques for formation pressure prediction while drilling,” Arab. J. Geosci., vol. 12, no. 18, 2019.
[10] E. Z. Naeini, S. Green, I. Russell-Hughes, and M. Rauch-Davies, “An integrated deep learning solution for petrophysics, pore pressure, and geomechanics property prediction,” Lead. Edge, vol. 38, no. 1, pp. 53–59, 2019.
[11] L. Hu et al., “A new pore pressure prediction method-back propagation artificial neural network,” Electron. J. Geotech. Eng., vol. 18 S, pp. 4093–4107, 2013.
[12] I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning. MIT Press, 2016.
[13] M. Kamber, Data Mining?: Concepts and Techniques, Second. Elsevier, 2016.
[14] C. Xiao, J. Ye, R. M. Esteves, and C. Rong, “Using Spearman’s correlation coefficients for exploratory data analysis on big dataset,” Concurr. Comput. Pract. Exp., vol. 28, no. 14, pp. 3866–3878, Sep. 2016.
[15] J. Han, M. Kamber, and J. Pei, Data Mining?: Concepts and Techniques, Third. Elsevier, 2012.
[16] L. Anastasia Kyrykovych, “Deep Neural Networks.” [Online]. Available: https://www.kdnuggets.com/2020/02/deep-neural-networks.html. [Accessed: 05-Jun-2020].
[17] F. Chollet, Deep Learning with Python. Manning Publications, 2017.
[18] C. Willmott and K. Matsuura, “Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance,” Clim. Res., vol. 30, pp. 79–82, Dec. 2005.