As a matter of general belief there does not exist a universal mathematical relationship between porosity and permeability. Common practice to draw relationship between these two parameters is usually graphical, for instance using picket plots or picking general trends in the core or well logs data. Trends picked are mostly explicit and subjective i.e. limited to well and cannot be extrapolated to full field if the degree of heterogeneity is significant. Core work presents a general rigorous procedure for deriving a linear mathematical model for predicting permeability given porosity and vice versa, using statistical methods i.e. regression analysis & probability distribution, is developed using appropriate core data.
Based upon the literature review to the best of our knowledge so far regression analysis presented between porosity & permeability is so shallow and is only limited up to drawing trends in core or well logs data. Whereas in this paper, application of standard regression analysis methods opted from statistics is discussed in more details. In this regards the utility of Shapiro-Wilk test, Quantile-Quantile (Q-Q) plot, Cook's distance and residual plots have been presented to deal with outliers which normally come from heterogeneity. These techniques ultimately help developing a linear mathematical model for prediction. Besides in any new oil & gas field for reserves calculation and production forecast usually average porosity and permeability is needed. However, for best estimates average values for porosity & permeability have been calculated using probability distribution techniques. Average values obtained are more reliable than those obtained using simple averaging methods thus relatively more precise estimates of reserves/resources and production forecast can be made.