Summary
Lateral propagation of foam in heterogeneous reservoirs, where pore geometries vary laterally, depends on the roles of pore geometries on the foam properties. In this paper, the pore attributes of 12 different rock samples were characterized in terms of porosity, permeability, pore shape, pore size, throat size, aspect ratio, coordination number, and log mean of surface relaxation times (T2LM). These were measured from gas porosimeter and permeameter, X-ray microcomputed tomography (CT)-basedpore-network models, thin-section photomicrographs, and nuclear magnetic resonance (NMR) surface relaxometry. The samples have a wide range of porosity: 12 to 29%; permeability: 1 to 5,000 md; average pore size: 3.7 to 9 µm; average throat size: 2.4 to 8 µm; average aspect ratio: 1 to 1.7; average coordination number: 2.6 to 5.2; and T2LM: 9.4 to 740 ms. Nitrogen foam flow experiments (without oil) were then conducted on each rock sample using a specialized coreflood apparatus. A graphical analysis of the coreflood data was used to estimate the total saturation of trapped foam (10 to 66%), flowing foam (3 to 14%), and apparent viscosity of foam (3.2 to 73 cp). Trapped foam saturation and apparent viscosity values were then correlated with each of the measured pore attributes. The results revealed that all pore attributes, except aspect ratio, have positive correlations with foam trapping and apparent viscosity. The best correlation with trapped foam saturation was obtained when the most influential pore attributes (pore size, throat size, aspect ratio, and coordination number) were combined into a single mathematical function. Foam apparent viscosity also appears to be mostly influenced by trapped foam saturation, permeability, and coordination number of pore systems. Trapping is also likely enhanced by the presence of fenestral or channel pores. Furthermore, the shape and angularity of pores seem to facilitate snap-off and trapping of foam, because rock samples with angular pores trapped the highest foam saturation compared with other samples with rounded and subrounded pores. It was also shown that the correlation between trapped foam saturation (and foam apparent viscosity) and the absolute permeability of porous media may reverse at some high-permeability values (greater than several darcies), when one or both of the following conditions exist: (1) The aspect ratio of a lower-permeability porous medium is lower than that of a higher-permeability porous medium, and (2) the coordination number of a lower-permeability porous medium is higher than that of a higher-permeability porous medium. Finally, T2LM showed a good correlation with foam trapping, making NMR logging a prospective tool for pre-evaluating foam performance in targeted reservoir sections.