No abstract

No abstract

Reservoirs such as Duvernay, Montney and Eagle Ford are segmented in different areas, ranging from predominantly dry gas regions, to wet gas and oil regions. Extensive research has focused on the application of enhanced recovery methods in the oil window of such reservoirs. In this paper, we discuss the application of enhanced recovery in the gas condensate window, with the objective to investigate the impact of diffusion on liquid dropout and vaporization on a matrix level. The Maxwell-Stefan equations were used to account for diffusion phenomena in the medium, and a phase behavior routine was implemented including nano-confinement effects. Numerical experiments were performed to evaluate the range of variability of recovery factors in a cyclic gas injection scenario. Methane was used as injection gas, and 1, 2 and 4 cyclic injection stages were modelled at the scale of a matrix block. Sensitivity was performed using a leaner and a richer gas composition, as well as two levels of permeability (50 and 100 nD). This allowed detailed investigation of time and location of occurrence of liquid dropout through saturation profile maps. Due to molecular partitioning, the phase envelope shifts as production proceeds, generating an accumulation of heavier hydrocarbons in the medium. Since injectivity is reduced in lower permeability media, injection pressure ramp up needs to be controlled to prevent condensate blockage. As a result, longer cycles are needed in the lower permeability case to achieve equivalent recovery. Liquid dropout is recurrent during production after each injection cycle, however, increasing the number of stages resulted in overall lower liquid saturation during subsequent production. Additionally, saturation profile maps indicate that the locus of condensate banks varies between each stage. As more injection stages are performed, a leaner gas remains in the vicinity of the fracture boundary and the condensate bank is formed further into the matrix block. Although more cycles improved recovery of heavier hydrocarbons, faster cycles resulted in lesser penetration of the injection gas into the porous medium. This behavior is more accentuated in the lower permeability cases. Nevertheless, recovery of heavier fractions is still higher compared to the primary production base case. Sensitivity studies will dictate the optimum number of stages for a fixed timeframe. In this work we use a combination of physics involved in flow in tight reservoirs to demonstrate how saturation profile maps can be used as a tool to improve enhanced recovery strategy.

Molecular diffusion is a transport mechanism often ignored in conventional, non-fractured multi-component petroleum upstream simulations due to the predominance of convection. In unconventional fractured reservoirs, diffusion plays a vital role in hydrocarbon production. A "shale" reservoir is characterized by thin, ultra-tight matrix blocks surrounded by natural or induced fractures. This creates conditions in which diffusion fluxes could be significant. In ultra-tight formations, convection is a slow process, and the presence of thin blocks surrounded by fractures increases the contact area, both of which favors diffusion. In this paper, we discuss the application of cyclic gas injection to enhance recovery in tight reservoirs in the gas condensate window. A fully implicit model is implemented with the objective to investigate the impact of diffusion on liquid dropout and vaporization on a matrix level. Diffusion fluxes are implemented considering a gradient in total chemical potential as driving force. Additionally, since capillary forces are significant in ultra-tight formations, phase equilibria calculations are modified to account for nano-confinement effects. Sensitivity is performed on matrix block size and injection gas composition (pure C1, a mixture of C1 and CO2, and a mixture of C1, C2 and C3), and the role of diffusion is evaluated for each scenario. As gas is injected, the composition of heavier hydrocarbon fractions in the gas phase significantly increases due to vaporization of condensate. Molecular diffusion helps to spread composition banks. As a result, liquid dropout is delayed during the subsequent production stages. Heavier fractions remain in the gas phase for longer periods, which ultimately enhances its recovery. In addition to that, retention of heavier fractions due to condensate dropout is intensified as the size of the matrix block increases. Longer matrix blocks result in lower swept length for the same number of cycles. As a result, liquid dropout occurs earlier because feed of gas at in-situ composition diffuses from the center of the matrix block towards the fracture boundary. We demonstrate that heavier components recovery is more affected by molecular diffusion than lighter components. Furthermore, it is observed that molecular diffusion strongly influences time and location of occurrence of liquid dropout in tight gas condensate reservoirs. Implementation of a rigorous model that includes convection, diffusion, adsorption and phase change allowed to investigate the commingling effects of different physics involved in enhanced recovery in unconventional reservoirs.

No abstract