'Flake-like Reflection' Carbonate Reservoir Appraisal: A Case Study in Tarim Basin

Abstract

Abstract 'Flake-like reflection' carbonate reservoirs in the Tarim Basin, a kind of secondary carbonate reservoirs formed by honeycomb vugs with complex pore shapes, have not been routinely explored like 'Bead-like reflection' reservoir due to its insufficient reflection understanding caused by rock-physics inaccuracy. Therefore, to build the relationship between seismic responses and reservoir properties as precise as possible, DEM-Gassmann model, which not only considers the influence of pore shapes on velocity but also meets the requirement of 'dilute' pores by incrementally adding pores, is employed in the first time to systematically model the pre- and post-stack reflections of 'Flake-like reflection' carbonate reservoir with different porosities and fluids in the Tarim Basin in this paper. The relationship between 'Flake-like reflection' reservoir properties and seismic responses is eagerly required for reservoir appraisal and well deployment, which will definitely improve reservoir appraisal and exploration efficiency in the Tarim Basin. Introduction The storage spaces of carbonate reservoir in the Tarim Basin are dominated by secondary pore types such as dissolution vugs, holes, and cracks (Figure 1) (Zhang et al, 2011). Many theoretical researches have proven that pore shapes can significantly influence the velocities of elastic waves propagating in rocks (Eshelby, 1957; Kuster and Toksöz, 1974; Xu and White, 1995; Yan et al., 2002; Sun et al., 2004). Empirical Vp-Vs relationships such as Castagna's (1985) mud-rock line, Han's relations (1986), and Greenberg-Castagna's (1992) relationship that ignore the effect of pore geometry are not applicable to carbonates (Xu and White, 1995; Wang et al., 2009). Wang et al. (2009) analyzed the practicality of three published models (Wyllie time average equation, Gassmann's equation, and Kuster-Toksöz model) on velocity prediction for carbonate reservoirs in the Tarim Basin, proving that Kuster-Toksöz model is the best one. Meanwhile, they also pointed out that Kuster-Toksöz model is a very high-frequency model, and is only limited to dilute concentrations of the pores. Based on effective medium theories of Norris et al. (1985), Berryman (1992) proposed a new differential effective medium (DEM) model. The matrix in this model begins as phase 1 and is updated at each step when a new increment of phase 2 (inclusions) is added, continuing the process till the expected proportion of the inclusions is reached. In this process, the former incremental inclusion (phase 2) and the matrix phase are taken as a new matrix, which make the inclusion added at each step is 'dilute'. Unfortunately, DEM is also a high frequency model as Kuster-Toksöz model. At high frequencies, there is no enough time for wave-induced pore pressure to equilibrate, so it is not applicable to consider the effect of pore fluids on elastic properties. DEM-Gassmann model which only employs DEM model (1992) to calculate the moduli of dry rock by modeling geometrical shapes of different secondary pores and meeting the requirement of 'dilute' pores through incrementally adding pores (both considering the effect of pore geometry, and eliminating the effect of velocity dispersion), and uses Gassmann's equation (1956) to calculate the moduli of saturated rocks (incorporating the effect of pore fluids), is proposed by Sam et al in 2012. Regarding the algorithm of 3D special pores, they used spheres, needles, and penny-shaped cracks to respectively represent dissolution vugs (small-sized caves), holes (needle-shaped), and cracks developed in the Tarim basin. These pores are incrementally added into the matrix of carbonate rocks in the order of volumes preferred, so that the inclusion added at each step is dilute.