The Influence of Sample Size and Permeability Distribution on Heterogeneity Measures

Abstract

Summary This study evaluates the influence of sample size on the Dykstra-Parsons and Lorenz measures of heterogeneity. Because either coefficient is <>determined for a reservoir from a finite number of data, only an estimate is made of the true coefficient. We show that, on average, the estimate is less than the true value. We give the relationship between estimate error and number of samples and show how significant errors may arise when too few data are used. The influence of the permeability distribution on heterogeneity measures is also studied. We show that a variety of distributions exhibiting different reservoir performance can have the same Dykstra-Parsons coefficient. We propose a heterogeneity measure that includes an indication of the permeability distribution. The new measure requires fewer data than the Dykstra-Parsons estimator. The relationships between the new measure and the Dykstra-Parsons and Lorenz coefficients are given.