Loading...

Permeability modeling using ANN and collocated cokriging

Zarei, A ; Sharif University of Technology | 2010

895 Viewed
  1. Type of Document: Article
  2. Publisher: Society of Petroleum Engineers , 2010
  3. Abstract:
  4. Obtaining a reliable reservoir permeability map that is consistent with all available data is of great important for reservoir engineers. However, there is not enough core from existing wells to estimate the reservoir permeability directly but Well log data are more widely available. This study aims to model permeability within the reservoir while there is no enough data. In particular, we use artificial neural networks to estimate permeability using four input logs of sonic, gravity, porosity and neutron logs in six existing wells. In order to eliminate the correlated data, we have done Principal Component Analysis on selected input logs. Collocated cokriging is considered as a valuable alternative to full cokriging while there is no adequate amount of primary variable to use univariate geostatistical methods like kriging and SGS. We used this in the framework of the sequential simulation to produce various realizations. As the permeability was overestimated in the shaley layers, a shale correction had to be performed on the estimated permeability logs. Then acoustic impedance with a good correlation to the permeability was used as covariable to do 3D modeling of permeability. The underlying statistics, histograms and cross validation confirms the degree of exactness of the presented model. © 2010, European Association of Geoscientists and Engineers
  5. Keywords:
  6. Acoustic impedance ; Engineering exhibitions ; Engineers ; Neural networks ; Neutron logging ; Petroleum engineering ; Principal component analysis ; Three dimensional ; Well logging ; 3-d modeling ; Artificial Neural Network ; Co-Kriging ; Collocated cokriging ; Correlated data ; Cross validation ; Geostatistical method ; Good correlations ; Kriging ; Neutron log ; Permeability modeling ; Reservoir engineers ; Reservoir permeability ; Sequential simulation ; Univariate ; Well log data ; Petroleum reservoir evaluation
  7. Source: 72nd European Association of Geoscientists and Engineers Conference and Exhibition 2010: A New Spring for Geoscience. Incorporating SPE EUROPEC 2010 ; Volume 5 , 2010 , Pages 3939-3943 ; 9781617386671 (ISBN)
  8. URL: http://earthdoc.eage.org/publication/publicationdetails/?publication=39672