Quantifying Bypassed Oil in the Vicinity of Discontinuous Shales

Abstract

Abstract We use a combination of analytic, experimental and numerical techniques to quantify flow around discontinuous shales and the extent to which they cause oil to be bypassed. We begin by using analytic solutions and well-characterised bead-pack experiments to investigate flow around a single, isolated shale during miscible (M=1, M=10) and immiscible displacements (M=10). We find that, regardless of the displacement, oil is initially bypassed downstream of the shale, in contrast to the conventional view that oil is bypassed upstream of a shale. This oil is ultimately recovered, although the time required depends upon the displacement (miscible/immiscible, M=1,10). We compare the analytic and experimental results with predictions obtained using conventional finite-difference simulation, and demonstrate that simulation can correctly model the flow around an isolated shale if suitably refined grids are used. We then use simulation to investigate the effect of multiple shales on flow patterns and the bypassing of oil. The grid resolution is important; simulations conducted on typical field-scale coarse grids generally overestimate the volume of bypassed oil in the vicinity of the shales. Introduction Most clastic reservoirs contain discontinuous shales which act as barriers or baffles to flow.1,2 These shales cannot be correlated between wells, and their likely location within the reservoir must be predicted using stochastic techniques conditioned to core and outcrop data.1,3,4,5 Their presence will clearly increase the tortuosity of fluid flow paths, and hence decrease the effective single-phase permeability of the reservoir (Fig. 1a).1,4–10 However, the extent to which they cause bypassing of oil has not been properly quantified; many papers have investigated single-phase flow around discontinuous shales, but only a few have investigated two-phase immiscible or adverse mobility ratio miscible flow. Studies of simple ‘generic’ shale models suggest that, during immiscible displacements, oil in the vicinity of a shale is initially bypassed by the displacing fluid, and subsequently drains from around the shale driven by viscous and gravitational forces (Fig. 1b).2,11,12 The initial bypassing of oil leads to early breakthrough of the displacing fluid, and if many shales are present, to disruption of the displacement front. However, if the timescale of oil drainage from around the shales is short compared to the timescale of reservoir production, the ultimate recovery of oil is unchanged.3 In contrast, studies of more realistic shale distributions13,14 suggest very little oil is bypassed except when shales are continuous over large areas of the bedding surfaces and are steeply inclined to the flow or if the displacement is highly unfavourable. These studies all rely on finite-difference simulations to model the flow, yet none properly establish the accuracy of the simulations; moreover, none determine quantitatively and in detail the true nature of the flow around a single shale. This is essential if simulators are to be used to predict volumes of bypassed oil that might be available to EOR/IOR techniques. In this paper, we begin by investigating flow around a single ‘isolated’ shale which is far from the influence of other shales and the reservoir boundaries. This simple system has not been previously investigated. We use a combination of analytic, experimental and numerical techniques, and investigate both miscible (M=1, M=10) and immiscible displacements. We assume that flow is viscous dominated, incompressible and perpendicular to the shale. These conditions maximize the bypassing of oil: if flow is oblique to the shale then there is a larger component of flow along the face of the shale, causing less oil to be bypassed,14 while gravity forces can reduce bypassing by driving oil drainage from around the shale.2 We find good agreement between the analytic, experimental and numerical results if suitably refined grids are used to simulate flow. This gives us confidence in the validity of the numerical solutions, and we use numerical techniques to investigate the effect of multiple shales on flow patterns and the bypassing of oil.