Hydraulic and Poroelastic Rock Properties From Oscillating Pore Pressure Experiments

Abstract

Knowledge of hydraulic and poroelastic properties is essential for simulating fluid flow in porous media. Accurate constraints on these properties have impacts on production forecasts and economics. In this paper we document an improvement of the oscillating pore pressure experiment by simultaneously measuring hydraulic and poroelastic properties of reservoir rocks. Measurements were carried out for four conventional reservoir rock quality samples at oscillation frequencies of 0.025-1 Hz and effective pressures of 3.5-62 MPa. Estimated permeability values decreased with increasing effective pressure and increased sharply above a frequency range of 0.3-0.4 Hz. We established that hydraulically measured storage capacities are overestimated by almost an order of magnitude when compared to elastically derived ones. Biot coefficient was estimated both from hydraulic and strain measurements, and comparison of two data sets reveals high uncertainty of the hydraulic specific storage measurements. We documented grain crushing and pore collapse event in a dolostone sample, observed as a permanent and drastic decrease of permeability and bulk modulus. We validated our method by detecting irreversible microstructural changes independently by hydraulic, elastic, X-ray microtomography ( CT) and nuclear magnetic resonance (NMR) measurements. Our approach can be used to constrain and to improve the estimation of specific storage and thus leads to better model inputs and forecasts.

Numerical Simulation of Oscillating Pore Pressure Experiments and Inversion for Permeability

“…Figure 3 unequivocally displays a frequency dependence of permeabilities, measured in all samples. While data collected by Fischer (1992) and Song and Renner (2006) show a decrease of permeabilities with increasing oscillation frequency, data by Bernabé et al (2004) and Hasanov et al (2019) demonstrate an opposite trend. Although the discussion of the frequency dependence of permeabilities measured by the oscillating pore pressure method is out of the scope of this paper, it is important to mention it.…”

“…First is to investigate how well the analytical model by Kranz et al (1990) describes the oscillating pore pressure experiment. Second is to take an alternative look at the frequency‐dependent amplitude ratio and phase data and reexamine apparent frequency dependence of permeabilities, previously reported by Hasanov et al (2019). Third, if the oscillatory pore pressure diffusion behaves according to Kranz et al (1990), then the inversion produces a set of best fit permeability and specific storage.…”

“…We demonstrate the influence of a sample's specific storage on amplitude ratio and phase by forward modeling the frequency response of these parameters for a fixed value of permeability. The simulation has been performed for Fox Hills 7 sample data (Hasanov et al, 2019) at 3.45 MPa effective stress, three values of specific storage ( β st =10 −9 , β st =10 −10 , and β st =10 −11 Pa −1 ) and three downstream reservoir volumes (100 cm 3 , 941 cm 3 (the actual downstream reservoir volumes used in experiments by Hasanov et al, 2017), and 1,500 cm 3 ). The chosen values of specific storage are representative of a typical sediment (sand), a sedimentary rock (sandstone), and a low‐permeability and low‐porosity igneous rock (granite), respectively, according to Table 3.1 from Wang (2000).…”

“…Permeabilities, calculated from a single‐point frequency measurements, are plotted as a function of oscillation frequencies. (a) Data from Hasanov et al (2019) used in this study. (b) Data from Fischer (1992), Bernabé et al (2004), and Song and Renner (2006).…”

“…We solve numerically the uncoupled pore pressure diffusivity equation with oscillatory boundary conditions for a given sample geometry to investigate the effect of the frequency of the pore pressure oscillation on amplitude ratio and phase. We also compare the results of numerical simulations with the analytical solution and representative data sets (Hasanov et al, 2019; Kranz et al, 1990).…”