The Role of Principal Component Analysis (PCA) in the Delineation of Lithofacies Based on Well Logs

Abstract

 A rich set of well logs were measured in a well in Ghadamis Basin, Libya. These data are used to compare the effectiveness of different methods used to delineate lithofacies boundaries in the borehole. Quantitative statistical lithology analysis, based on six logs, is found to be the most effective.

It utilizes geological information, particularly the knowledge of mineral types occurring in the formations. It provides volume fractions of minerals including porosity, volumes of shale and ferroan minerals. The depth interval can be divided into segments of porous reservoir rock, sandwich-type development with alternating sand and shale laminae, and impermeable shale.

The statistical method of principal component analysis (PCA) is applied for lithofacies analysis is in the same well, lithofacies categories are selected on the crossplots of the first principal component (PC1) vs. the second principal component (PC2). Formation boundaries are delineated based on lithofacies analysis.

Two PCAs were applied: one from the six logs, also used for lithology interpretation, the other from two inputs: gamma ray and conductivity log inputs. Lithofacies were delineated based on both of the PCAs and compared with the results of detailed lithology interpretation. PCA is also used for volumetric estimation of porosity, volumes of shale, and ferroan minerals.

pdf