Prediction of gas-oil capillary pressure of carbonate rock using pore network modeling

This study suggests a frame work to predict the capillary pressure curve of the carbonate rocks using a pore network simulation. A novel pore network manipulation algorithm is introduced to match three curves and predict only one curve (air-oil). This implies that the matching process is much more complicated. In return, more valid results are obtained and the problem of non-uniqueness of statistical PNM approach is mitigated. Finally, a collection of the most significant parameters governing capillary pressure are tabulated. Two plugs with air-mercury, air-oil and water-oil drainage and water-oil imbibition capillary pressure tests are used. Water-oil drainage and imbibition and air-mercury drainage are matched with the same pore network. The pore network is constructed based on a statistical distribution of the network elements. Then, properties of air and oil are input to the pore network to estimate air-oil capillary pressure. Matching three capillary pressure curves using a network is a complicated task to do, especially when no 3D image of the plugs is available. On the other hand, it is evident by the acceptable results that the error of prediction is moderated using the suggested method. This estimation technic is a new application of pore network modeling that can be used for reservoirs with no 3D image of plugs and limited SCAL data.