1887

Abstract

Summary

A common issue in geological modeling is to determine facies along each well. It is easy in cored wells, but difficult in uncored wells where only electric logs are available. The issue is usually solved by defining “electrofacies,” calculated from logs and calibrated with core data to ensure their geological consistency.

The most common clustering approaches generally consider all the log measurements as equivalent, and rarely care about the spatial correlation between the data points. However, accounting for the spatial relationship between samples in clustering will improve the realism and the geological consistency of the electrofacies.

This paper presents a case study made with a clustering algorithm which accounts for the spatial relationship between samples, allowing a balance of the influence of the different sources of information: log responses, lithological description and sample locations.

Working with a search neighbourhood to define the pairs of samples to be used for the hierarchical clustering, combined with the influence of coordinates, allows one to manage a large range of non-stationary effects. By comparing the statistics calculated on the classification results to the sedimentological model or to Vertical Proportion Curves computed from cored wells, optimal input parameters for the clustering algorithm can be determined.

Loading

Article metrics loading...

/content/papers/10.3997/2214-4609.201900708
2019-06-03
2020-03-28
Loading full text...

Full text loading...

References

  1. ThomasRomary, JacquesRivoirard, JacquesDeraisme, CristianQuinones, XavierFreulon
    (2012). Domaining by clustering multivariate geostatistical data. Ninth International Geostatistics Congress, 2012, Oslo, Norway. pp.455–466.
    [Google Scholar]
  2. Romary, T., Ors, F., Rivoirard, J., & Deraisme, J.
    (2015). Unsupervised classification of multivariate geostatistical data: Two algorithms. Computers & Geosciences, 85, pp. 96–103.
    [Google Scholar]
  3. Friedman, J., Hastie, T., & Tibshirani, R.
    (2001). The elements of statistical learning. Vol. 1. No. 10. New York, NY, USA: Springer series in statistics.
    [Google Scholar]
http://instance.metastore.ingenta.com/content/papers/10.3997/2214-4609.201900708
Loading
/content/papers/10.3997/2214-4609.201900708
Loading

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