Effective Pressure and Microstructure Control on Resistivity Formation Factor and Seismic Waves Velocities

摘要

In clay-rich tight gas sand, conventional resistivity methods yield inconsistent saturation estimates, and simple velocity models such as the Biot-Gassmann do not accurately predict the saturation induced velocity variations in these formations. This paper presents an experimental study designed to better understand the electrical and acoustic properties of clay-rich tight gas sands. A comprehensive set of measurements including porosity - permeability, mineralogy, mercury capillary pressure and SEM imaging were conducted on six tight gas sands cores saturated with 25,000 ppm NaCl brine. To complete the experiments, resistivity, compressional and shear wave velocities were measured on the same tight gas sands as a function of hydrostatic confining pressure (500 psi to 5000 psi) and pore pressure (500 psi to 3000 psi). During these experiments, the time necessary for pore pressure equilibrium was determined to be at least three hours. The tight gas sands studied have porosities ranging from 4.5% to 8.3% and permeabilities between 18 μd and 82 μd. Their mineralogy is dominated by the presence of quartz with clay content as high as 28 wgt%. The SEM imaging shows the presence of a significant number of low aspect ratio pores and authigenic microcracks. The formation factor and seismic velocities data were used to evaluate the effective pressure law for each of these petrophysical properties. The effective pressure dependency for the formation factor and the seismic velocities observed are in opposition to each other. The analysis of the effective pressure coefficient n for the formation factor showed a general increase from a value generally close to 1, by 31% over a differential pressure range of 500 to 2500 psi. This same coefficient n generally equals about 0.92 for low differential pressure but decreased by about 30% for compressional and shear wave velocities. Using simple considerations, we established that a change in n values by 30% can induce an error in pore pressure prediction of 15%.