Hybrid method for porosity classification in carbonate formations

Recently the methods of intelligent computing (IC) have been applied for the interpretation of well log data. This is due to the necessity to process well logs in the situations when complete information about them cannot be obtained. In this case, hybrid methods based on statistical analysis, fuzzy logic and evolutionary algorithms could be very useful. This paper presents such hybrid analysis of logging data of the wells from the Cantarell Oil Complex, in the Zonda of Campeche, Mexico. Different methods, such as principal component analysis, factor analysis, fuzzy classification and evolutionary optimization are used for analysis of the structure of porosity space given by primary, cavernous and micro-fractures porosity classes. Comparison and analysis of the obtained results show that IC methods can substantially compensate for the absence of exact information without losing the precision of data analysis and at the same time decrease the costs of well logging.