Quality Checking and Prediction of Capillary Pressure:Comparison of Alternative Models with Laboratory Measurements
NGUYEN, Vy Le, YAN, Zhiwei
Engineering Honors - 2015
Australian School of Petroleum
University of Adelaide
Capillary pressure is an important parameter for assessing and modelling fluid flow through porous media. Traditionally, capillary pressure is measured in the laboratory by techniques such as mercury injection, porous plate, and centrifuge. However, laboratory measurements are commonly expensive and time consuming. In many cases, capillary pressure relationships are not available for every reservoir zone of interest. Therefore, a modelling approach is typically used to first quality check laboratory data and to fill in data gaps.
The first section of the project gives a literature review of capillary pressure, laboratory experiments and a historic account of alternative capillary pressure models. The main focus of this study is to experiment with a recently developed capillary pressure model, the Modified Purcell model (Goda and Behrenbruch, 2011), which incorporates four parameters: maximum capillary pressure, irreducible water saturation, entry pressure and the curvature of relationships. The modelling approach involves the translation of a relationship into model space. Confined by endpoints, the relationship plots as a straight line in the model space, a relationship between Capillary Pressure Index (CPI) and Effective Saturation Group ESG). The study investigated two datasets from Australian fields: Bayu-Undan and Perseus South, where data was generated by different laboratory techniques, porous plate and centrifuge respectively.
The study concludes that the new model works well for characterizing diverse plugs: homogeneous, bimodal, high permeability streak, broad pore size distribution. This modelling approach also gives insight to rock quality based on the location of linear relationships in model space. For prediction of capillary pressure, model entry pressure has been used, typically lower than the lab entry pressure, where the latter is not well defined, in part due to spontaneous imbibition, and also due to finite pressure increments taken by the lab.