Simulation and interpretation of formation-tester measurements acquired in the presence of mud-filtrate invasion and geomechanical deformation.

摘要

Wireline formation testers are widely used to measure in-situ fluid pressure, to retrieve reservoir fluid samples, and to estimate formation mobility. However, formation-tester measurements are invariably influenced by mud-filtrate invasion due to drilling overbalance pressure, thereby affecting the acquisition of uncontaminated fluid samples and the estimation of in-situ petrophysical properties. Moreover, in cases of stress-sensitive formations, rock mechanical deformation may take place due to the combined effects of in-situ stress, wellbore stress imposed by mud overbalance, and wellbore pressure exerted by the formation tester itself. The latter deformation causes near-borehole perturbations of porosity and permeability that are evidenced by pressure transients measured during build-up and shut-in stages of formation testing, especially when using dual-packer pressure probes. If unaccounted for, such perturbations can also bias the estimation of in-situ fluid and petrophysical properties. Conversely, the detection and quantification of elastic mechanical deformation effects on measured pressure transients can be used to infer the underlying rock elastic and petrophysical properties of the stressed formation.;The purpose of this dissertation is twofold: (a) to quantify the relative effects of mud-filtrate invasion and geomechanical deformation on pressure-transient measurements acquired with dual-packer formation testers, with special emphasis on the appraisal of near-borehole porosity and permeability enhancement due to elastic mechanical deformation, and (b) to develop a new method to estimate elastic and petrophysical properties of rock formations from dual-packer pressure transients acquired in mechanically deformable rocks.;Numerical simulations of mud-filtrate invasion are performed with an axial-symmetric two-phase (water-oil) method that enforces the specific boundary and source conditions of a wellbore that penetrates horizontal layers. Simulations are performed in a cylindrical system of coordinates using finite differences together with an implicit-pressure, explicit-saturation time-marching approach that also incorporates the dynamic conditions of immiscible mudcake growth due to filtration of solids at the wellbore. Laboratory experiments are conducted to further study pressure transients due to formation testing in the presence of invasion with water-base mud. Experiments include the effects of both mud circulation and mudcake on pressure-transient measurements and are performed on a variety of rock-core samples. Measurements are successfully validated with both the developed simulator and a commercial simulator, thereby lending credence to the assumed model of dynamic solid filtration.;The developed mud-filtrate fluid-flow simulator is coupled with a finite-element code that assumes 2D axial-symmetric linear elasticity to quantify geomechanical deformation. Coupling of mechanical deformation with variations of porosity and permeability assumes a staggered-in-time, iteratively coupled volumetric model. We assume a dual-packer formation tester to quantify elastic deformation effects in stress-sensitive formations as a preamble to estimating in-situ elastic and petrophysical properties. It is shown that near-wellbore spatial variations of porosity and permeability due to mechanical deformation can bias the corresponding pressure-transient measurements acquired with the dual-packer formation-tester. The degree of biasing depends on the rigidity of the stressed formation.;Finally, we develop a method to estimate in-situ petrophysical and elastic rock properties from pressure-transient measurements acquired with formation-testers in mechanically deformable rocks. Petrophysical and elastic properties will change in both time and space depending on the time evolution of the conditions that influence mechanical deformation. We use a commercial reservoir simulator to calculate pressure transients due to fluid pumpout in the presence of both invasion and mechanical deformation. A pre-stressed initial condition due to mud overbalance is assumed with incremental deformation due to surface force applied by the packers or probes, and active flow imposed by the formation-tester. In so doing, we consider pressure data sets acquired with both flow and observation probes during draw-down and build-up periods. For cases where a-priori information can be sufficiently constrained, our estimation method provides reliable and accurate estimates of petrophysical and elastic properties in the presence of moderate levels of random noise.