Defining irreducible water saturation using global characteristics envelopes and novel correlations

Abstract

Irreducible water saturation plays a significant role in estimating hydrocarbon initially-in-place and petroleum recovery. Yet, laboratory measurements for determining irreducible water saturation take considerable time and money. For this reason available data may not cover all requirements, giving rise to the practise of using correlations to fill in gaps. Described in this paper are the reasons for irreducible water saturation being an elusive parameter that not only depends on pore structure characteristics but also the type of experiment and laboratory procedures, as well as changing plug conditions during experimentation. This paper reviews traditional methods, as well as recent and novel approaches to quality assure laboratory data and for correlating irreducible water saturation for prediction. To gain insight into the dependence of irreducible water saturation on detailed pore structure characteristics, most notably grain size and sorting, the usefulness of global characteristics envelopes is explored (Behrenbruch and Biniwale, 2005). In this multidimensional plot, irreducible water saturation is plotted against porosity, permeability, hydraulic radius, porosity group, flow zone indicator (grain size) and sorting, giving an insightful overview of the interdependence of parameters. The second part of this paper compares novel correlations with commonly used correlations. Traditional and more recent correlations are covered, from simple correlations versus the logarithm of permeability to more sophisticated approaches using more variables, including porosity and others. Most notably, it is shown that an approach of correlating irreducible water saturation with grain size (or flow zone indicator [FZI]) and sorting shows great promise. Data from two Australian fields are used to demonstrate the methodology, showing a significant increase in fitting accuracy. This approach may eventually lead to a universal correlation.