An automatic identification method of log diagenetic facies and its application in Sulige area, Ordos Basin

Key words:  petroleum reservoirs  lithology  well logging  diagenetic facies  diagenesis  probabilistic neural network  clay mineral

Abstract:

Lithology was used as a unit to extracting log feature and identifying diagenetic facies, which could remove the lithology influence on the well-logging informations and control the development types of diagenetic facies in the identification unit.The clay content ratio was used to reflect acidic or alkaline of pore fluid, and the identification accuracy of diagenetic facies was improved. Based on the final effect of diagenetic facies on the development of pore, the conception of log diagenetic facies was given. By using probabilistic neural network based on Bayes optimal classification, the identification ability of diagenetic facies with different logs was compared. The identification coincidence rate of log diagenetic facies is 81％ by this method in He 8 section of Sulige area, Ordos Basin. And the gas test results are general gas reservoirs or gas and water reservoirs.The poor gas reservoirs barely exist in the area where corrosion facies develop widely.