Reservoir fluid type identification from petrophysical logs using pattern recognition methods

Identifying the type and distribution of reservoir fluids is one of the main things on well logging logs and well testing. Several methods have been proposed to identify the type of reservoir fluids that in general، it can be divided into two groups of methods; direct (e. g. well testing) and indirect methods (e. g. seismic and log interpretation). Petrophysical logs due to their high resolution and more conformity are more frequently used than seismic data. This study aims to identify reservoir fluid types in PLs، based on 3 classes of oil، oil-water and water، in carbon reservoir. Suggested method applies wavelet decomposition as well as classification and was applied to PLs in five wells of an oil field in southwestern Iran. Eventually، obtained results have been evaluated by well testing responses.