Predicting Porosity through Fuzzy Logic from Well Log Data

Porosity is one of the most important characteristics for modeling reservoir. In recent years, some new methods for estimation have been introduced, which are more applicable and accurate than old methods. Fuzzy logic has shown reliable results in petroleum modeling area for describing reservoir characteristics. In this study, a Sugeno fuzzy model has been formulated to predict porosity. In order to select the number of membership function, subtractive clustering method was utilized through Gaussian membership functions. Another technique for predicting porosity was multiple linear regression to compare with fuzzy logic technique. Results indicated that correlation between real value from core data and the predicted value by fuzzy logic was more accurate than multiple linear regression technique in this scope.