Permeability Prediction from Petrophysical Logs Using Fuzzy Logic in Clastic-Carbonate of Asmari Formation

Abstract

Aboozar field is located approximately 72 km west of Khark Island in offshore Iran. Main reservoir of this field is an Oligo-Miocene stratum called Ghar which is a member of Asmari. Permeability plays important role in reservoir flow and always has much difficulty to be defined and modeled. This property is directly measured based on core analysis and drill stem test. This paper presents the use of fuzzy logic molding to prediction permeability from wireline log data in clastic-carbonate reservoirs of Asmari formation. Fuzzy logic allows quantifying qualitative information and modeling complex not linear relation. In essence, the problems consist in finding some explicit relation between log and core in well. After training the fuzzy model with subtractive clustering in a given well, permeability prediction were made for other layers in the same well. In subtractive clustering a one and quick way shall be selected for training which based on this a Takagi-Sugeno-Kang method can be generated. The number of membership functions and laws doesn’t need to be defined because inferring system is able to define those by using clustering input-output data. These predictions were in excellent agreement with permeability values obtained from core.