ANALYSIS OF AN UNSUPERVISED CLUSTERING ALGORITHM FOR LITHOFACIES CLASSIFICATION IN RIGA FIELD, NIGER DELTA by Olisa Benson Akinbode and Oluniyi Taiwo Samuel, Scientific Research Journal (Scirj), Volume XIII, Issue IV, April 2025 Edition

Scirj Volume XIII, Issue IV, April 2025 Edition
ISSN: 2201-2796

ANALYSIS OF AN UNSUPERVISED CLUSTERING ALGORITHM FOR LITHOFACIES CLASSIFICATION IN RIGA FIELD, NIGER DELTA

Olisa Benson Akinbode and Oluniyi Taiwo Samuel

Abstract: The study determined the optimum number of clusters that effectively capture lithological variations in subsurface formations and also emphasizes the application of these clustering algorithms for lithofacies classification. The k-means clustering algorithm was used to uncover hidden patterns in well logs in the Niger Delta RIGA field. This model successfully classified the data into 4 distinct clusters, revealing correlations with depth and gamma ray (GR) measurements. The developed clustering model was able to automatically classify the dataset into useful clusters, these clusters, when matched with depth, generated useful lithologies in the well RIGA-1 and RIGA-2. There are 4 clusters having 3 very visible clusters similar to Continental sands, Marginal marine sandstones and Shale. These mostly tie up with the changes in the logging measurements, decrease in Gamma Ray (GR) from around 4900m to 6100m, 7700 to 7600, 8500 - 9700 aligns with the blue clusters in Well 1 and Well 2.

Reference this Paper: ANALYSIS OF AN UNSUPERVISED CLUSTERING ALGORITHM FOR LITHOFACIES CLASSIFICATION IN RIGA FIELD, NIGER DELTA by Olisa Benson Akinbode and Oluniyi Taiwo Samuel published at: "Scientific Research Journal (Scirj), Volume XIII, Issue IV, April 2025 Edition, Page 19-31 ".

Search Terms: unsupervised clustering algorithm, lithofacies classification

[Read Research Paper] [Full Screen]