Experimental Investigation on Permeability and Porosity Hysteresis in Low-Permeability Formations

Abstract

Summary Permeability and porosity decrease with increasing net stress in consolidated and unconsolidated porous media is a well-known phenomenon to petroleum and geomechanics engineers. Conversely, permeability and porosity are observed to increase when net stress decreases, however typically follow a different path, this discrepancy is known as hysteresis. The trend of permeability and porosity hysteresis is a signature of porous medium that depends on several physical and mechanical properties. Understanding permeability and porosity hysteresis plays a significant role in production strategies of hydrocarbon reservoirs. Hysteresis effect on production strategies can be even more important in very low permeability reservoirs such as tight sandstone, tight carbonate, and shale formations. The reason is that the stress change associated with permeability and porosity hysteresis can affect adsorption/desorption and diffusion transport mechanisms which are among the main driving mechanisms in low or ultra-low permeability reservoirs. In this study, permeability and porosity hysteresis of nano, micro, and milli-Darcy core samples are measured for a wide range of net stress and the results are correlated with pore structures observed by Field Emission Scanning Electron Microscopy (FE-SEM) images. The nano and micro-Darcy core samples are from the Niobrara, Bakken, Three Forks, and Eagle Ford formations. The milli-Darcy core samples are from Middle East carbonate, Indiana Limestone, and Torrey Buff Sandstone formations. Bakken, Three Forks, and Middle East carbonate are core samples from oil producing reservoirs, whereas others are from outcrop. Our experimental observations show that: (a) compared to steady state method, the unsteady state permeability measurement used here can shorten the time required to measure permeability of nano and micro-Darcy cores without significant measurement error; (b) stress dependency of permeability and porosity and their hysteresis during loading and unloading of outcrop cores were observed to be very small compared to those of reservoir cores; (c) generally, the stress dependency and hysteresis of permeability and porosity was observed to be inversely related permeability and porosity at initial stress conditions; hence, these characteristics are more pronounced and important in organic rich shale reservoirs; (d) porosity decrease with stress and its hysteresis were observed to be much smaller than that of permeability; (e) our results generally follow the experimental and theoretical relationship between cubic root of permeability and logarithmic of net stress reported by many researchers; and (f) an increase in permeability with temperature (and decrease in permeability with hysteresis during temperature unloading) is observed in organic rich mudrocks. This is due to dilation, diffusion, and adsorption effects of organic rich pores. Hence, a temperature correction factor is required to estimate the exact in-situ permeability of organic rich mudrocks. Finally, we would like to comment that understanding of permeability and porosity dependency on stress and their hysteresis information values could be used to optimize production of low-permeability formations such that the irreversible permeability loss (formation damage) could be minimized by early fluid injection (pressure maintenance) to minimize rapid stress increase due to fluid withdrawal from stimulated reservoir volume. Also, these permeability and porosity hysteresis information could be used in proper designing of multi-stage hydraulic fracture and re-fracturing design of organic-rich low-permeability formations.