Investigation of density inversion induced by gas charges into oil reservoirs using diffusion equations

Abstract

A reservoir crude oil consists of dissolved gas, liquid and dissolved solid (asphaltenes). It is well known that asphaltene solubility in oil decreases with an increase in solution gas content (gas/oil ratio, GOR). For example, a late gas charge into a reservoir crude oil over geologic time can cause asphaltene instability and phase separation to form a tar mat. This tar mat is often spatially separated by significant distances from the location of the asphaltene instability, thus requiring asphaltene migration in porous media at rates greatly exceeding diffusive velocities. Therefore, a dynamic mechanism is proposed to account for these known observations. Diffusive entry of gas from a gas cap into an oil column gives rise to chemical potential gradients (diffusion driving forces). The cross-term effect in diffusion makes asphaltenes move in the opposite direction to the concentration gradient in a certain region, which is referred to as the gas expulsion effect. On the other hand, the diagonal-term effect in diffusion moves asphaltenes along the concentration gradient. Consequently, both competing effects result in asphaltene concentration fingering at a certain location and thus creates fluid density inversion. In turn, this fluid density inversion in the oil column leads to gravity currents (gravitational instability – diffusion induced convection) that enable the migration of asphaltenes over large distances in porous media over geologic time.A one dimension (1-D) diffusive model with a moving boundary is developed in this paper coupled with the FHZ EOS(Flory-Huggins-Zuo equation of state ) and applied to three component (gas, maltene and asphaltene) systems. The gas/oil contact moves up with time due to the swelling effect of a gas charge into a crude oil reservoir. Density inversion can be produced mainly by the cross-term effect in diffusion (i.e., the impact of the presence of charging gas on the chemical potential of asphaltenes or gas expulsion) in a relatively wide range of conditions. Parameter sensitivity analyses demonstrate that significant fluid density inversion is generated at the conditions close to the asphaltene phase instability boundary, which is not surprising because the cross-term effect in diffusion becomes significant at those conditions. In addition, the diffusion model can be used to identify asphaltene phase instability directly. Application of this diffusive model to oilfield cases is in progress.