Page 186 - Machine Learning for Subsurface Characterization
P. 186

Dimensionality reduction and clustering techniques Chapter  6 161


             the drastic permeability contrast between fractures and matrix, the injected fluid
             prefers to flow through open fractures leaving behind large portion of oil in the
             matrix. Alharthy et al. [16] conclude that miscible mixing and solvent extraction
             near the fracture-matrix region is the primary oil mobilization mechanism of
             miscible light-hydrocarbon injection. Jin et al. [14] and Dadmohammadi et al.
             [19] showed that both viscous and molecular flow regimes can exist in
             organic-rich shales and ultra-tight formations; viscous flow exists in fractures,
             while molecular flow and viscous flow can simultaneously exist in matrix.
                Shales have dominant pore size at nanoscale resulting in pore-confinement
             effects on the MMP. It is well known that the flow behavior in nanoscale is
             significantly different from that in larger scale [20, 21]. Fluid properties in
             nanopores differ greatly from bulk fluid, which means conventional
             calculation may not accurately describe the miscibility. Light-hydrocarbon
             injection in shales occurs at lower MMP because the critical temperature and
             pressure for the injected light hydrocarbon and reservoir oil decrease in
             nanopores [22]. Several methods have been proposed to calculate the MMP in
             nanopores [23, 24]. An alternation of MMP due to pore-confinement effect
             may result in miscibility variations in pores as a function of pore size, which
             is adequately accounted for in this study.
                Volume content of bitumen, which is a nonproducible organic matter
             deposited in the pores, is another important factor governing the EOR
             efficiency. Due to large molecular size of bitumen and small pore throat size
             in the formation, the existence of bitumen clogs pores and inhibits the oil
             flow. Furthermore, bitumen increases kerogen swelling giving rise to smaller
             and more complicated pore structure [25]. Another factor influencing the
             EOR efficiency is the water content. At high water saturation, oil is
             surrounded by water in the pore; therefore the sweep efficiency of miscible
             displacement is reduced [26]. In this chapter, we take both bitumen and
             water as negative factors for the EOR using light-hydrocarbon injection.
                Natural or induced fractures may pose two contrasting effects on the oil
             recovery with light-hydrocarbon injection. The presence of fracture enhances
             the mobilization of oil in matrix by molecular diffusion. Nonetheless,
             connected fracture system may result in low sweep efficiency. Natural or
             induced fractures increase the contact area between matrix oil and injected
             fluid and thus enhance the diffusion process near the fracture-matrix interface.
             The density, orientation, and size of fractures determine the producibility of
             shales [27], which has been ignored in this study.
                Kerogen content and pore wettability also have significant impact on EOR
             efficiency when using light hydrocarbon. Experimental investigation showed
             that kerogen content is the most important factor affecting miscible gas
             injection performance [14]. Pores in shale formations can be classified as
             pores in organic matter and those in nonorganic matter. Kerogen is
             predominantly composed of micropores and mesopores that are oil wet and
             exhibit complex pore structure. Consequently, it is harder to displace
             hydrocarbon from pores in organic matter.
   181   182   183   184   185   186   187   188   189   190   191