Correlation of capillary pressure data and the development of the widely used J(S) correlating function are reviewed. The pore structures of core samples, which give the same (in pore structures of core samples, which give the same (in practice closely similar) J(S) curves, are described as practice closely similar) J(S) curves, are described as operationally similar. Similarity of J(S) curves can he recognized by parallelism of semi-log plots of capillary pressure vs. saturation. Curves so identified were grouped to pressure vs. saturation. Curves so identified were grouped to obtain a pore structure factor based on J(S) curve shapes rather than determining Purcell lithology factors for individual formations. The modified Purcell method gave improved correlation between measured and predicted permeabilities.
A general solution of the Thomeer model in terms of the tangent angle for the Pc(S) curve is presented which facilitates tests of correlations between permeability and selected features of capillary pressure curves. Single point correlation methods based upon the tangent as introduced by Swanson for= 45, are shown to hold for a wide range of saturation. Curve methods which consider all of the curve such as suggested by Purcell or any selected parts can he tested at will. The method is designed for semi-empirical investigation of a very large database of core analysis results which include porosity, permeability, and capillary pressure by mercury porosity, permeability, and capillary pressure by mercury injection.
There is a widespread interest in characterization of porous media through interpretation of capillary pressure, Pc, porous media through interpretation of capillary pressure, Pc, relationships. In the petroleum industry, mercury injection into initially evacuated dry rock is commonly used to measure Pc curves needed for reservoir engineering purposes. Pc curves needed for reservoir engineering purposes. Particular attention has been given to development of Particular attention has been given to development of correlations which permit estimation of permeability from mercury injection measurements on drill cuttings and fragmented core samples m order to obtain estimates of permeability for formation evaluation purposes. permeability for formation evaluation purposes. Mercury injection curves are basically primary drainage curves (drainage from 100% initial saturation of the wetting phase, which is essentially a vacuum and will he referred to as phase, which is essentially a vacuum and will he referred to as such). These curves are commonly interpreted, using a parallel bundle of tubes model, to obtain pore sire parallel bundle of tubes model, to obtain pore sire distributions. The model has many obvious weaknesses. Pores in sedimentary rocks are neither straight nor circular and are far from uniform in cross-section. An early approach to allowing for nonuniformity of cross-sections involved cutting and rejoining tubes of various sizes.
Considerable attention has been given to modelling porous media as networks. Pores within the network have a pore body size and are connected to other pores by pore throats of smaller size, which control entry into the pore body. The numbers of pore throats for each pore body is determined by the coordination number of the network. The manner in which pore Sizes and pore volumes are distributed over the network pore Sizes and pore volumes are distributed over the network (correlated pore throat to pore body sizes and/or correlations of pore sizes between neighboring pores) have also been modelled. All of the above features and, just as importantly, the choice of rules for displacement in the network, strongly contribute to the shapes of capillary pressure curves, and also affect absolute and relative permeabilities and other properties calculated for a network. Network models provide a useful guide to causes of variation in shape of capillary pressure curves according to the network design and the rules used to calculate displacement behavior for the network.
A more direct approach to relating capillary pressure curves to a pore size that takes network properties into account distribution was proposed by Mason; it is based on interpretation of drainage and imbibition scanning curves.