Improving Pore Pressure Estimation by Rock-physics-based Mineral Substitution

Abstract:

Pore pressure is an important factor controlling rock properties and fluid flows in sedimentary basin and subsurface reservoirs. Existing methods of pore pressure prediction usually relate pore pressure to compaction, and utilize acoustic velocity as a proxy for porosity to estimate pore pressure. Nevertheless, in addition to porosity, velocity is also influenced by other factors such as mineralogy, fluid, and pore structure. At shallower depth of the earth’s crust, the change of elastic moduli of minerals with pressure is negligible and pressure dependency of velocity is only due to variations of fluid compressibility, porosity and pore structure with pressure. Pressure-independent variation of lithology with depth, however, can severely affect the accuracy of pore pressure prediction as changes in velocity caused by mineral variation may be interpreted as changes in pore pressure if velocity alone is used such as in Eaton’s method. To minimize the effects of pressure-independent mineral variation on pore pressure in such methods, we propose a rock-physics method to develop a new velocity profile that retains only the pressure-dependent components. The method uses Gassmann equation first to extract fluid contributions and obtain dry bulk modulus, then a rock physics model is further used to separate contribution of mineral to dry bulk/shear modulus from those of porosity and pore structure. Effects on dry bulk/shear modulus from mineral variation can be calculated using the rock physics model by substituting the original depth-varying mineral composition with a single mineral. A new velocity profile can be obtained which preserves the true pressure influences on fluid, porosity, and pore structure. The proposed method has been successfully applied to well log data from a shale formation interbedded with limestones in Sichuan basin, China. Pore pressure estimated using the corrected velocity profile agrees with mud weight data better than using the original Eaton’s method.