Alternative modelling approaches for the estimation of irreducible water saturation: Australian hydrocarbon basins

Abstract

Estimation of original oil in-place requires precise knowledge of certain reservoir parameters, such as porosity and irreducible water saturation. In addition to its importance in defining oil in-place, irreducible water saturation, Swir, is a key parameter in evaluating multi-phase flow, for example relative permeability. Residual oil saturation, the target of tertiary recovery, is also a function of Swir. The primary method of determining Swir is by conducting capillary pressure experiments, requiring considerable resources and long time periods, with the consequence of a limited number of core plug evaluations for a particular reservoir. For the above-mentioned reasons, the estimation of Swir with mathematical models may be attractive. The study reported here uses artificial neural networks and a semi-empirical equation that is calibrated using a Genetic Algorithm to determine Swir. The performance of these models is compared with other, conventional models, demonstrating the superior performance of the proposed Swir prediction models. All models are calibrated with data from Australian hydrocarbon basins, but the outlined approach is expected to be applicable to other data sets also. © 2006 Elsevier B.V. All rights reserved.