A committee machine approach for predicting permeability from well log data: a case study from a heterogeneous carbonate reservoir, Balal oil Field, Persian Gulf

Abstract

Permeability prediction problem has been examined using several methods such as empirical formulas, regression analysis and intelligent systems especially neural networks and fuzzy logic. This study proposes an improved and novel model for predicting permeability from conventional well log data. The methodology is integration of empirical formulas, multiple regression and neuro-fuzzy in a committee machine. A committee machine, a new type of neural network, has a parallel structure in which each of the applied methods (experts) has a weight coefficient showing its contribution in overall prediction. The optimal combination of the weights is obtained by a genetic algorithm. The method is illustrated using a case study from a heterogeneous Upper Jurassic carbonate reservoir in Balal oil Field, Persian Gulf. For this purpose, one hundred fifty-one samples from the intervals comprising core and well log data were clustered into eighty-one training sets and seventy testing sets to evaluate the validity of the models developed. The results of this study show that the genetic algorithm optimized committee machine has provided more accurate results than each of individual experts used