Pacific Section AAPG Convention

Datapages, Inc.Print this page

Deriving Pseudo-Capillary Pressure Curves From Standard Core Analysis Data In Heavy Oil Reservoirs And Their Use In Estimation Of Original Sw


Original water saturation is needed for reservoir studies, yet can be difficult to determine accurately in older fields. Capillary pressure (Cp) data can be used to calculate original Sw, but for many reservoirs suitable core material is not available or special core analysis was not done. This paper presents a method to use standard PKS core data in heavy oil reservoirs to construct pseudo-Cp curves for application in a petrophysical model to calculate original Sw. Since heavy oil does not flush significantly from the core, the saturations from PKS analysis are representative of reservoir saturations. To apply this method, all valid core data from the reservoir are compiled into a single dataset. For each core point, the height above the free fluid interface (oil/water contact) is calculated. Then the set of core analyses is sorted into permeability groups (bins), the ranges of which are determined from the available data. Within each permeability bin, core Sw (X axis) is plotted against height above the FFL (Y axis). Several trial groupings may be required to determine the best way to bin the data. Since most reservoirs have undergone some depletion, the data within each permeability group will probably be spread across a range of Sw values. However, some data points are likely to be close to original Sw. A curve can be fitted to the low-Sw edge of the data cloud for each permeability group, using the predicted behavior of Cp curves to shape them where data may be sparse. A permeability-referenced set of pseudo-Cp curves can thus be derived and used to construct a petrophyscial model. These curves can also supplement laboratory measurements that do not capture the full range of permeability variation in the reservoir. Using log-derived permeability and height above the OWC for each log depth level, original Sw can be estimated. This method bypasses problems with variable Rw values, heat-affected resistivity curves, and unknown amounts of reservoir depletion. With porosity computed from modern logs, OOIP can be estimated, and these data are suitable for reservoir modeling. In many fields, the method is applicable to old e-logs since a permeability transform can often be developed based on short normal curves. For this case, porosity can be determined by first computing clay volume from the SP, GR, or other clay indicators. The porosity is then obtained by a transform which is a function of clay volume and maximum porosity (PHImax) in a clean sand.