Compositional modeling of retrograde gas-condensate reservoirs in multi-mechanistic flow domains

A multi-mechanistic flow environment is the result of the combined action of a Darcian flow component (the macroscopic flow of the phase due to pressure gradients) and a Fickian-like or diffusive flow component (diffusive flow due to molecular concentration gradients) taking place in a hydrocarbon reservoir. The present work presents the framework needed for the assessment of the impact of multi-mechanistic flow on systems where complex fluid behavior - such as that of retrograde gas-condensate fluids - requires the implementation of compositional reservoir simulators. Due to the complex fluid behavior nature of gas-condensate fluids, a fully-implicit (IMPISC-type) compositional model is implemented and the model is used for the study of the isothermal depletion of naturally fractured retrograde gas reservoirs. In these systems, especially those tight systems with very low permeability (k < 0.1 md), bulk fluid flow as predicted by Darcy's law might not take place despite the presence of large pressure gradients. The use of an effective diffusion coefficient in the gas phase - which accounts for the combined effect of the different diffusion mechanisms that could take place in a porous medium - and its relative contribution to fluid recovery is discussed. The compositional tracking capabilities of the model are tested and the conditions where Fickian flow can be the major player in recovery predictions and considerably overcome the flow impairment to gas flow posed by the eventual appearance of a condensate barrier - typical of gas-condensate systems - are investigated. Finally, a mapping that defines different domains where multi-mechanistic flow can be expected in compositional simulators of retrograde gas-condensate reservoirs is presented.