Discrepancy of simulating snap-off processes in 2D and 3D pore-throats

3D pore-throats with uniform thickness were commonly used to explore snap-off processes in experimental studies, and always simplified as 2D areal models in simulation of snap-off processes. However, 3D effects of capillary force, viscous force and wall cannot be neglected. We assume that snap-off processes in 2D and 3D porethroats may exist discrepancy. Snap-off processes in both 2D and 3D pore-throats were simulated with various capillary numbers Ca, viscosity ratios mu r and size ratios M of throat and pore. The results showed: (1) snap-off was more prone to occur with the same M in 3D pore-throats due to the larger capillary pressure difference; (2) the increase of M inhibited occurrence of snap-off, and the inhibition effect of M was more obvious in 2D porethroats; (3) snap-off only occurred with M < 7/35 in 2D pore-throats, which was significant different from the Roof theory (M < 1/2). These findings indicate that 2D areal models are not suitable to simplify the 3D porethroat in simulating snap-off processes.