Abstract
The aim of this study is to predict permeability from well log data for a heterogeneous carbonate reservoir in the upper part of the Sarvak Formation (mid-Cenomanian to early Turonian) in an Iranian oil field. The permeability is a crucial parameter for reservoir modeling which is difficult to measure directly. Direct measurements of permeability are obtained form core plugs, if available. Continuous permeability values, however, need to be predcited indirectly from independent data. Three different methods have been tested for estimating the permeability: 1) permeability from effective porosity 2) multilinear regression, and 3) fuzzy logic. The two latter methods utilize raw well logs (gamma ray, density, neutron and sonic) to predict permeability. Core plug measurements have been used to validate the predictions.
Results from the study show that fuzzy logic yields better results than the two other methods. The multilinear regression is unable to represent the dynamic range measured on the core plugs (overestimated in low values and underestimated in high value). The permeability predicted from the effective porosity model is more or less similar to multilinear regression model, but relatively narrower. This is however to be expected as these models both indicates narrower ranges than fuzzy logic, because they are trying to fit with average values.
