Abstract
The numerical simulation of heavy oil steamfloods has generally been conducted assuming that the oil is non-volatile. Reservoir simulation has traditionally ignored compositional effect s due to heat and steam and assumed that the hydrocarbon phase is non-volatile. This is equivalent to assuming that the equilibrium ratios, K-values, are zero. In order to properly model the mechanism of steamflooding, however, compositional effects need to be taken into account. In this study, laboratory data including distillation, vapor pressure, steam distillation and viscosity measurements, along with a commercial PVT simulator are used to tune equation-of-state (EOS) and viscosity parameters to properly model the PVT properties of the oil. The Peng-Robinson equation-of-state (PR-EOS) was used for all phase behavior calculations. Viscosity as a function of temperature and composition was modeled with the Pedersen correlation for heavy oils. Once a tuned equation-of-state, compositional fluid description was developed for the heavy oil, one-dimensional numerical simulations of the steamflooding process were performed. These simulations demonstrated the utility of the equation-of-state approach. In addition, it was concluded that compositional effects are essential for the proper modeling of low residual oil saturations seen in the field and the formation of an in-situ solvent bank at the steam-oil interface.
Lolley, Christopher Scott (1995). Compositional changes in heavy oil steamflood simulators. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1995 -THESIS -L646.