• Economic and Environmental Geology
• /
• v.44 no.2
• /
• pp.161-170
• /
• 2011

The multiphase flow simulator, CHEMPS, was developed based on the fractional flow approach reported in the petroleum engineering literature considering fully three phase flow in physically and chemically heterogeneous media. It is a extension of MPS developed by Suk and Yeh (2008) to include the effect of wettability on the migration of NAPL. The fractional flow approach employs water, total liquid saturation and total pressure as the primary variables. Most existing models are limited to two-phase flow and specific boundary conditions when considering physically heterogeneous media. In addition, these models focused mainly on the water-wet media. However, in a real system, variations in wettability between water-wet and oil-wet media often occur. Furthermore, the wetting of porous media by oil can be heterogeneous, or fractional, rather than uniform due to the heterogeneous nature of the subsurface media and the factors that affect the wettability. Therefore, in this study, the chemically heterogeneous media considering fractional wettability as well as physically heterogeneous media were simulated using CHEMPS. In addition, the general boundary conditions were considered to be a combination of two types of boundaries of individual phases, flux-type and Dirichlet type boundaries.

• The Journal of Engineering Geology
• /
• v.19 no.1
• /
• pp.81-88
• /
• 2009

Two dimensional Monte-Carlo simulations of a non-reactive solute plume in isotropic porous media which are physically and chemically heterogeneous are conducted to determine the variations of moment. Retardation factors of 1, 2 and 5 are given to ascertain how the second moments are changed as adsorption increased. Retarded longitudinal second spatial moment, ${Z_{11}}^{'R}(t',l')$, increased during the transport process and as the dimensionless lengths of line plume source, $l_2'$, increased. ${Z_{11}}^{'R}(t',l')$ decreased as the retardation factors increased, and the simulated moments fit well to the first-order analytical results. Retarded longitudinal plume centroid variance, ${Z_{11}}^{'R}(t',l')$, decreased as the dimensionless lengths of line plume source, $l_2'$, increased and as the retardation factor increased. The result indicates that the uncertainty about the plume center decreased, and the ergodic condition for the second spatial moments is far from reaching. Simulated longitudinal one particle displacement covariance, ${Z_{11}}^{'R}(t')$, well consistent with the first-order analytical results for the three degrees of retardation factors of 1, 2 and 5 respectively. It is, consequently, concluded that the retarded longitudinal second moments could be produced by stochastic simulation, and that the first-order analytical results definitely provides very close values of the longitudinal retarded moments.