Fluvial facies reservoir productivity prediction method based on principal component analysis and artificial neural network

Abstract

It is difficult to forecast the well productivity because of the complexity of vertical and horizontal developments in fluvial facies reservoir. This paper proposes a method based on Principal Component Analysis and Artificial Neural Network to predict well productivity of fluvial facies reservoir. The method summarizes the statistical reservoir factors and engineering factors that affect the well productivity, extracts information by applying the principal component analysis method and approximates arbitrary functions of the neural network to realize an accurate and efficient prediction on the fluvial facies reservoir well productivity. This method provides an effective way for forecasting the productivity of fluvial facies reservoir which is affected by multi-factors and complex mechanism. The study result shows that this method is a practical, effective, accurate and indirect productivity forecast method and is suitable for field application.

Keywords

Fluvial facies reservoir / Productivity forecast / Principal component analysis / Artificial neural network

References

[1] Tan Chengqian, Ma Narui, Su Chao, Model and method for oil and gas productivity prediction of reservoir, J. Earth Sci. Environ. 26 (2) (2004) 42-45.

[2] Fan Ting’en, Li Weixin, Wang Zhihong, et al., Aresearch of prediction and description techniques for fluvial reservoir in Bozhong 34 area, Bohai sea, China Offshore Oil Gas 1 (2006) 13-17.

[3] Liu Zongyan, Shi Yanwei, Research of reservoir productivity prediction by well logging, World Well Logging Technol. 22 (5) (2007) 35-44.

[4] M.L. Cheng, Productivity prediction from well logs in variable grain size reservoir Cretaceous Qishn Formation, Republic of Yemen, Log. Anal. 40 (1) (1999) 24-34.

[5] Kuang Jianchao, Tong Xiaohua, Yang Weining, The method for predicting productivety of dense clastic reservoir in west Sichuan depression, J. Mineral. Petrol. 14 (4) (1994) 85-91.

[6] M. Shahab, Virtual Intelligent Applications in Petroleum Engineering, SPE 62415, 2000.

[7] D.P. Hampson, J.S. Schuelke, J.A. Quirein, Use of multiattribute transforms to predict log properties from seismic data, Geophysics 66 (1) (2001) 220-236.

[8] O. Christian, J. Michael, The Definitive Identification of Candidate Wells for Refracturing, SPE 84211, 2003.

[9] S.L. Jong, Reservoir properties determination using fuzzy logic and neural network from well data in offshore Korea, J. Pet. Sci. Eng. 49 (3-4) (2005) 182-192.

[10] W. William, How Artificial Intelligence Methods Can Forecast Oil Production, SPE 75143, 2002.

[11] Zhang Qinghong (Master thesis),Design and Implementation of Marine Geological Disaster Forecasting System Based on the Principal Component Analysis, vol. 12, Qingdao University of Science and Technology, 2010.

[12] Li Shixin, Wen Jian, Shao Xiaohou, et al., Sediment yield model for small watersheds based on coupling of partial least square regression and artificial neural network, J. Hohai Univ. Sci. 38 (2) (2010) 149-152.

[13] Yang Shuning, Pattern Recognition and Intelligent Computing:Matlab Technology, Electronic Industry Press, 2008.

[14] Tong Kaijun, Shan Yuming, Liao Tequan, et al., Statistics-based reservoirproductivity ranking prediction by well logging, Nat. Gas Explor. Dev. 31 (2) (2008) 40-43.

[15] H. Cinco, F.G. Miller, J. Ramey, et al., Unsteady-State Pressure Distribution Created by a Directionally Drilled Well, SPE 5131, 1975.

[16] Luo Xianbo, Zhao Chunming, Wu Haiyan, et al., Determination of the interference coefficient among the commingled production layers in offshore oilfield, Pet. Geol. Oilfield Dev. Daqing 31 (5) (2012) 102-104.

[17] Ge Hanquan, Shi Zejin, Ren Zaiqing, Prediction of the reservoir parameters based on the coupling of neural network model with partial least square method, J. Chengdu Univ. Technol. Technol. Ed. 34 (6) (2007) 618-620.