AAPG Bulletin, Vol.105, No.4, 785-807, 2021
Application of principal component analysis on chemical data for reservoir correlation: A case study from Cretaceous carbonate sedimentary rocks, Saudi Arabia
A new method is presented to identify chemostratigraphic zones by applying principal component (PC) analysis and to interpret geological boundaries based on PC curves. The PC data were calculated using inductively coupled plasma-optical emission spectrometry and inductively coupled plasma-mass spectrometry, with data acquired for 50 chemical elements in the range Na-U. In the present study, a total of 832 core and cuttings samples were analyzed from four wells penetrating the Cretaceous Wasia Formation, southeastern Saudi Arabia. The conventional workflow involves plotting profiles for elements and elemental ratios, with 250 to 300 profiles produced for each study section. With so many variables to consider, it is a challenging and time-consuming process to produce chemostratigraphic frameworks, requiring expert knowledge of chemostratigraphy. By using only a few PC curves, as opposed to element or ratio profiles, it is easier for nonexperts to identify correlative boundaries. The PC parameters reflect changes in more than one group of elements and summarize geological changes. For example, in this study, PC2 parameter values relate to changes in depositional redox conditions. A close association was revealed between boundaries identified using PC and boundaries from an existing conventional chemostratigraphic study on the same wells. For example, PC zone E1-3 relates to the conventional zones C3-1b. The fact that some PC and conventional chemostratigraphic boundaries do not always coincide (e.g., conventional boundaries C2-1:C2-2 and C2-2:C2-3 exist within PC zone E2-1) means that the integrated correlation scheme is of higher resolution if these techniques were employed in isolation.