عنوان مقاله [English]
Total organic carbon (TOC) is one of the important parameters for the evaluation of the hydrocarbon generation potential of source rocks. The measurement of this parameter requires conducting geochemical analysis on cutting samples, which is expensive and time consuming. In general, organic rich rocks are characterized by higher porosity, higher sonic transit time, lower density, higher gamma ray, and higher resistivity compared to other rocks. In this study, the linear and non-linear genetic algorithm models were used to estimate TOC from petrophysical data for the Kazhdumi, Gurpi, and Pabdeh source rocks in Ahwaz oilfield. The linear genetic algorithm model provided more reliable and acceptable results than the non-linear model. The genetic algorithm solutions for fitting coefficients to TOC equations were compared to a regression analysis. Performance analysis based on MSE and correlation coefficient indicates the higher performance of the intelligently derived equations in comparison to the statistical regression analysis. In the next step of the study, a cluster analysis technique was utilized for the classification of the estimated TOC log and the identification of geochemical zones. In the light of the acceptable results of the GA model, source rocks were classified into the organic-rich and organic-lean facies by using a cluster analysis method.