Efficient and Accurate Estimation of Permeability and Permeability Anisotropy FromStraddle-Packer Formation Tester Measurements Using the Physics of Two-PhaseImmiscible Flow and Invasion

摘要

This paper describes the successful application of a novelmethodology to estimate permeability and permeabilityanisotropy from transient measurements of pressure acquiredwith a wireline straddle-packer formation tester. Unlikestandard algorithms used for the interpretation of formationtester measurements, the methodology developed in this paperrigorously incorporates the physics of two-phase immiscibleflow as well as the process of mudcake buildup and invasion.The effects of supercharge, skin, and convective transport ofsalt are also explicitly included in the estimation algorithm.An efficient 2D (cylindrical coordinates) implicit-pressureexplicit-saturation finite-difference algorithm is used tosimulate both the process of invasion and the pressuremeasurements acquired with the straddle-packer formationtester. Initial conditions for the simulation of formation testermeasurements are determined by the spatial distributions ofpressure and fluid saturation resulting from mud-filtrateinvasion. Inversion is performed with a rigorous Levenberg-Marquardtnonlinear minimization algorithm. Sensitivityanalyses are conducted to assess non-uniqueness and theimpact of assumptions on rock-fluid and mud properties suchas fluid viscosity, capillary pressure, relative permeability,mudcake growth, and time of invasion.Applications of the estimation algorithm to noisy syntheticmeasurements include homogeneous, anisotropic, layered, andmulti-layered formations, for both low- and high-porosityrocks. We also study the effect of unaccounted impermeablebed boundaries on the inversion results. For cases where a-prioriinformation can be sufficiently constrained, ourinversion methodology provides reliable and accurateestimates of permeability and permeability anisotropy.