Facies Modeling using Back-propagation Neural Networks In Southern Pars Field

Document Type : Research Paper

Abstract

Determination of different facies is one of the most important and fundamental tasks of geological and engineering characterization of reservoir rocks. The neural network method is one of the new techniques used in identification of facies. The objective of the present study was to identify and measure different facies of Southern Pars gas and oil fields (Iran) using back-propagation neural networks in order to develop static and dynamic models. Modeling was carried out using three different techniques. Also network parameters were optimized in order to improve the network performance including number of layers and neurons, transfer function, training algorithm, dividing and performance functions. The results indicate that the back-propagation neural network is a powerful method for identification and modeling of the facies.

[1
] Satter A. & Thakur G.C., Integrated petroleum reservoir management, PennWell Publishing Company,
1994.
[2]
Bhatt A.,Reservoir properties from well logs using neural networks, Norwegian University of Science and Technology, chap. 6, 2002.
[3]
White A.C., Molnar D., Aminian K., Mohaghegh S., Ameri S. & Esposito P.,The application of ANN for zone identification in a complex reservoir”, SPE Paper 30977, SPE Eastern Regional Conference & Exhibition, Morgantown, West Virginia, USA, 1995.
[4]
Farmer R.G. & Adams S.J., Facies recognition using neural networks”, New Zealand Petroleum Conference Proceeding, 1998.
[5]
Schoenicke O., Alawi S.M., Bemani A.S., Kalam M.Z. & Varlet X.L.,Preliminary studies on using artificial neural networks to predict sedimentary facies of the Permo-Carboniferous Glacigenic Al Khlata formation”, Oman, SPE Paper 53260, SPE Middle East Oil Show, Bahrain, 1999.
[6]
Bhatt A. & Helle H.B., “Determination of facies from well logs using modular neural networks”, Petroleum Geosciences, Vol. 8, pp. 217-228, 2002.
[7]
Lianshuang Q. & Timothy R.C., Neural network prediction of carbonate lithofacies from well logs, Big Bow and Sand Arroyo Creek fields, Southwest Kansas”, Computer & Geosciences, Vol. 32, pp. 947-964, 2005.
[8]
Dubois M.K., Bohling G.C. & Chakrabarti S.,Comparison of four approaches to a rock facies classification problem”, Computer & Geosciences, Vol. 33, pp. 599-617, 2006.
[9] Maiti S., Tiwari R.K. & Kumpel H.,
Neural network modeling and classification of lithofacies using well log data: a case study KTB borehole site”, Geophysics J. Int., Vol. 169, pp. 733-746, 2007.
[10] منهاج م.ب.، مبانی شبکه‌های عصبی، دانشگاه صنعتی امیرکبیر (پلی تکنیک تهران)، 1387.
[
11] White A.C., Molnar D., Aminian K., Mohaghegh S., Ameri S. & Esposito P., The application of ANN for zone identification in a complex reservoir”,SPE Paper 30977, SPE Eastern Regional Conference & Exhibition, Morgantown, West Virginia, USA, 1995.
[12] Demuth H., Beal M. & Hagan M., Neural network toolbox 6 (Users Guide), The MathWorksTM, 1992-2009.
[13] Malki H.A. & Anwar M.S., “Determination of lithofacies from well logs using unsupervised neural network model”, The Technology Interface, The Electronic Journal for Engineering Technology, Vol. 5, No. 1, pp. 947-964, 2003
.
[14] حسنی‌پاک ع.ا.، شرف‌الدین م.، تحلیل داده‌های اکتشافی، انتشارات دانشگاه تهران، 1380.