Prediction of Irreducible Water Saturation: Comparison of Alternative Models with Laboratory MeasurementsLong Hoang Nguyen
The Van Nguyen
Engineering Honours 2013
Australian School of Petroleum
University of Adelaide
Irreducible water saturation (Swir) is one of the fundamental parameters used to determine characteristics of reservoirs, to determine, for example, estimates of original hydrocarbon-in-place and recovery factors. Estimates of Swir may be obtained from laboratory methods or with less accuracy from well logs. Common laboratory methods include the use of centrifuge and porous plate methods.
This study first concentrates on summarizing laboratory experiments and procedures before describing results of the main investigation, the study of various models that may be used to correlate laboratory data. Models included in this study involve classical methods, namely logk vs. Swir, Tixier, Timur and Coates, as well as more recent methods related in most cases to the theory of Kozeny (1927) and Carman (1938), or C-K. More recently (Ameafule et al., 1993) introduced the so called Flow Zone Indicator (FZI) to the C-K formulation, which has been successfully used to characterize geologic depositional environments and also in modeling of fluid saturation profiles. The research described in this report is based on a large and diverse SCAL dataset from Australian oil and gas fields, both onshore and offshore.
Various alternative model results for predicting Swir were compared before concentrating on FZI models. The latter were also used to gage the accuracy of different laboratory methods and core cleaning techniques. It has been found that Swir is controlled by distinctive factors: pore structure, mineralogy, wettability and laboratory testing conditions. This comparative study also reveals the effect on Swir of non-optimal laboratory methods, concluding that centrifuge methods in combination with cool solvent cleaning tend to result in the most realistic Swir determination.