The objective of this research was examination of machine learning algorithms in combination with a priori geological information applicability for automatical facies distribution from wireline logs problem. This study was based on data from Field M located in Western Siberia which can be characterized by complex geology making results of examination reliable.

During the project different classification algorithms were evaluated to find the most appropriate one for automatical facies interpretation task. Classifiers were trained and tested on data from Field M, produced results were compared by different metrics.

At the next step chosen classifier (Random Forest algorithm) was used for comparison of two machine learning approaches - standard and hierarchical. The latter uses a priori geological information, in this study facies zonation map acted as such information. Application of this expert knowledge during automatical facies distribution allows separation of the initial data set into subsets to simplify classification task and improve prediction accuracy.

Finally, developed algorithm was performed on the entire oilfield including more than 700 wells to justify its applicability for real industry problems.

Previously mentioned steps were conducted with aid of originally developed Python script which can be integrated into any software environment to automate facies interpretation process.


Article metrics loading...

Loading full text...

Full text loading...


  1. Dubois, M. K.
    [2007]. Comparison of four approaches to a rock facies classification problem.
    [Google Scholar]
  2. Fernandez-Delgado, M.
    [2014]. Do we need hundreds of classifiers to solve Real World classification problems?
    [Google Scholar]
  3. Hall, B.
    [2016]. Facies classification using Machine Learning.
    [Google Scholar]
  4. Tengelidi, D.
    [2014]. Fourier spectrums clustering for automated facies recognition of Field Y.
    [Google Scholar]

Data & Media loading...

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error