Back Propagation and Hidden Weight Optimization Algorithms Neural Network for Permeability Estimation from Well-Logs Data in Shaly Sandstone Petroleum Reservoirs: Application to Algerian Sahara

  • Leila AliouaneEmail author
  • Sid-Ali Ouadfeul
  • Amar Boudella
Part of the Advances in Science, Technology & Innovation book series (ASTI)


In this paper, we present an inexpensive approach based on a multilayer neural network, using two different algorithms to estimate permeability in petroleum reservoirs from well-logs data. In a supervised learning, the Back propagation (BP) and Hidden weight optimization (HWO) are tested in order to determine the best algorithm for better permeability predictions. The application to real data has been realized in the Algerian Sahara, exploiting data of several petrophysical parameters of Triassic reservoirs of two wells. The data of the first well are used to train the neural network machine as a pilot well, while the second well data were used for generalization to predict permeability. The obtained results are compared with permeability from core data.


Shaly sandstone reservoir Neural network Permeability Training algorithm 


  1. 1.
    Tiab, D., Donaldson, E.C.: Petrophysics: Theory and Practice of Measuring Reservoir Rock and Fluid Transport Properties, 2nd edn. Elsevier (2004)Google Scholar
  2. 2.
    Ellis, D.V., Singer, J.M.: Well Logging for Earth Scientists, 2nd edn. Springer (2007)Google Scholar
  3. 3.
    Ouadfeul, S., Aliouane, L.: Total organic carbon estimation in shale-gas reservoirs using seismic genetic inversion with an example from the Barnett Shale. Lead. Edge. 35(9), 790–794 (2016)Google Scholar
  4. 4.
    Balan, B., Mohaghegh, S., Ameri, S.: State-of-the-art in permeability determination from well log data: part 1—a comparative study, model development. In: Eastern Regional Conference & Exhibition, West Virginia, U.S.A. SPE, vol. 30978 (1995)Google Scholar
  5. 5.
    Aliouane, L., Ouadfeul, S., Djarfour, N., Boudella, A.: Permeability prediction using artificial neural networks. A comparative study between back propagation and Levenberg–Marquardt learning algorithms. In: LNESS. Springer (2013)Google Scholar
  6. 6.
    Aliouane, L., Ouadfeul, S.-A., Djarfour, N., Boudella, A.: Petrophysical parameters estimation from well-logs data using multilayer perceptron and radial basis function neural networks. In: Huang, T., Zeng, Z., Li, C., Leung, C.S. (eds.) ICONIP 2012, Part V. LNCS, vol. 7667, pp. 730–736. Springer, Heidelberg (2012)Google Scholar
  7. 7.
    Aqil, M., Kita, I., Yano, A., Nishiyama, S.: Neural networks for real time catchment flow modeling and prediction. Water Res. Manag. 21, 1781–1796 (2007)Google Scholar
  8. 8.
    Yu, C., Manry, M.T.: A modified hidden optimization algorithm for feedforward neural networks. In: Conference Record of the Thirty-Sixth Asilomar Conference on Signals, Systems and Computers, vol. 01, issue 2, pp. 1034–1038 (2002)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Leila Aliouane
    • 1
    Email author
  • Sid-Ali Ouadfeul
    • 2
  • Amar Boudella
    • 3
  1. 1.Laboratoire Physique de la Terre (LABOPHYT), Faculté des Hydrocarbures et de la ChimieUniversité M’hamed Bougara de BoumerdesBoumerdesAlgeria
  2. 2.University of Khemis MilianaKhemis MilianaAlgeria
  3. 3.Geophysics DepartmentUSTHBAlgiersAlgeria

Personalised recommendations