Secondary porosity index effect on improving permeability estimation from petrophysical logs utilizing artificial intelligent approaches

Permeability estimation using core data and petrophysical logs is a conventional approachwhich bears high uncertainty especially in carbonate reservoir characterization. In essence,the problem consists not only due to coring expenses rate, but also ambiguity in findingproper explicit log correlation to core data. Moreover, utilizing the correlated formula inwells without core data can pose errors. In this research the permeability was estimated from conventional petrophysical logs and it was calibrated with permeability obtained from core lab experiments. Applied intelligent systems are the matter of this research for permeability values estimation. To construct permeability estimation model, three techniques have been applied including conventional ANN, the Gonzalez, and Hambalek fuzzy logic techniques. These methods were applied in two wells drilled in Surmeh reservoir in Balal field to establish ANN and to derive a relation between core and well. The models were applied in control well in order to check the reliability and capability of models to estimate representative permeability value. The result showed however three foresaid techniques for permeability estimation were successful the secondary porosity distributed the correlation due to its reduction effect on permeability so that they were not interconnected. Therefore this effect was omitted using secondary porosity index in which the permeability estimation were improved and were estimated close to core value.