Abstract
Relative permeability and capillary pressure are the governing parameters that characterize multiphase fluid flow in porous media for diverse natural and industrial applications, including surface water infiltration into the ground, CO_{2} sequestration, and hydrocarbon enhanced recovery. Although the drastic effects of deformation of porous media on singlephase fluid flow have been well established, the stress dependency of flow in multiphase systems is not yet fully explored. Here, stressdependent relative permeability and capillary pressure are studied in a waterwet carbonate specimen both analytically using fractal and poroelasticity theory and experimentally on the microscale and macroscales by means of Xray computed microtomography and isothermal isotropic triaxial core flooding cell, respectively. Our core flooding program using water/N_{2} phases shows a systematic decrease in the irreducible water saturation and gas relative permeability in response to an increase in effective stress. Intuitively, a leftward shift of the intersection point of water/gas relative permeability curves is interpreted as an increased affinity of the rock to the gas phase. Using a microscale proxy model, we identify a leftward shift in pore size distribution and closure of microchannels to be responsible for the abovementioned observations. These findings prove the crucial impact of effective stressinduced pore deformation on multiphase flow properties of rock, which are missing from the current characterizations of multiphase flow mechanisms in porous media.
Similar content being viewed by others
Introduction
An accurate representation of the variation in relative permeability and capillary pressure with phase saturation is a crucial requirement for understanding multiphase fluid flow in hydrocarbon/geothermal reservoirs^{1,2,3,4}, vadose zone water transport^{5}, and carbon dioxide storage projects^{6,7}. Many researchers have studied the multiphase flow properties of porous media experimentally at both porescale and corescales with a focus on fluidfluid and fluidsolid interfacial interactions^{8,9,10,11,12,13,14,15,16,17}. Additionally, several empirical models have been developed to fit curves on the experimental data to reproduce multiphase flow properties^{18,19,20,21,22,23}. Recent analytical developments on this front have led researchers to suggest a fractal description of pore sizes in porous rocks^{24}. The investigation of flows in porous media is also an important research field within the numerical community. Complex flows in porous media across scales and in the presence of catalytic or other chemical reactions have been recently numerically performed in^{25,26,27,28} wherein the authors presented novel and efficient numerical approaches to study flows in fractured media. Although diverse in methodology and material type, these experimental, analytical, and numerical studies have routinely treated the porous media as a stressindependent solid with zero solidsolid mechanical interaction. This restriction in the theoretical treatment persists despite the limited, but increasing, body of experimental^{29,30,31,32,33,34} and analytical^{35,36,37} evidence of changes to multiphase flow properties with effective stressinduced deformation. Complicated physical behaviors and contradictory findings continue to challenge our understanding of stressdependent multiphase flow at porescale and corescale^{36,37}.
Fluid production or injection in subsurface porous media will locally change pore pressures and insitu stresses^{38,39,40}. These stress changes will, in turn, lead to pore deformation in response to the pore pressure/stress coupling effect, for which we use the term “geomechanics” to generally describe this process. The earliest theory to account this coupling effect of soils was introduced by Terzaghi^{41}. However, the linear theory of poroelasticity was first borne from the pioneering work of Biot^{42} and is broadly understood from corescale and porescale experimental studies^{43,44,45}. Several models have been presented to relate relative permeability and capillary pressure curves to the pore size distribution in porous media^{18,19,20}. Typically, few parameters are required to model these curves; these include the irreducible wetting phase (Sw_{ir}) and/or residual nonwetting phase (Snw_{r}) saturations, endpoint relative permeability (k_{r−max}), and one more parameter (λ), which defines any curvature in the dependency of relative permeability on the pore size distribution^{19}. A further enhancement of the model was achieved by treating the pore space as a bundle of capillary tubes using fractal (D_{f}) and tortuosity fractal (D_{t}) dimensional parameters^{46,47,48,49,50}. The idea of fractal distribution of capillary tubes was then used to solve classic relative permeability and capillary pressure equations^{24}.
Despite an abundance of experimental studies of relative permeability and capillary pressure from both porescale and corescale studies, the impact of pore deformation induced by effective stress changes on multiphase flow properties remains poorly explored^{33}. For conditions of increasing effective stress, existing experimental evidences^{29,32} reveal contradictory shifts in the relative permeability curves. Recognition of this contradiction has resulted in several authors conducting experiments to tackle the challenges in corescale experiments^{33,34} and porescale physical models^{44}, independently. However, a comprehensive study of multiphase flow properties (e.g. relative permeability and capillary pressure) in both corescale and porescale under a wide range of effective stress conditions has yet to be undertaken.
In this study, we use Xray computed microtomography to quantify the structure and shape of the pores, together with the pore size distribution and two metrics (fractal dimension and degree of anisotropy), of an initially waterwet Indiana limestone specimen at atmospheric pressure. Then, through a series of coreflooding experiments using water/N_{2} pair phases, the stressdependent porosity, absolute permeability, relative permeability, and capillary pressure of the same specimen are measured under triaxial isotropic effective stress (10 MPa to 30 MPa) and isothermal (40 °C) conditions. Using the measured stressdependent porosity and pore strain of the core, we reconstruct 3D porescale proxy models of the sample at 10, 20, and 30 MPa effective stress conditions. This approach provides a simultaneous investigation of the pore and/or throat shape alteration and pore size distribution indices (e.g. fractal dimension) under variable stress conditions. Additionally, an analytical model is developed based on the fractal theory to reproduce and interpolate the stressdependent relative permeability and capillary pressure curves. We find that increasing the effective stress results in a leftward shift in pore size distribution, closure of microchannels, and modification to the flow path tortuosity. These changes are responsible for a dramatic upward shift in capillary pressure curve and a decrease in gas relative permeability. Executing experiments under raising effective stress conditions and constant gas flow rates, we show that increasing applied capillary pressure at the porescale due to an increase in gas pressure provides additional energy for the gas phase to invade smaller pore channels, leading to a subsequent decrease in the irreducible water saturation. These findings emphasize the complexity of the physical processes responsible for geomechanical controls on multiphase flow properties in porous media. Based on this research, we discuss the significant role of the stressdependent relative permeability, capillary pressure, and irreducible wetting phase saturation for some interesting applications in engineering and natural process.
StressDependent Flow Properties from CoreScale Experiments
A set of singlephase (water) and twophase (water/N_{2}) core flooding experiments were conducted on a waterwet Indiana limestone core in a highpressure triaxial cell under a range of isotropic effective confining stress conditions from 0 to 30 MPa and a constant temperature at 40 °C. The equipment^{34} allowed for accurate measurement of porosity, absolute permeability, relative permeability, and capillary pressure at several different effective confining stress conditions.
First, the confined carbonate core inside the cell was fully saturated with deionized water using the procedure given in the Materials and Methods. Then, the required confining stress inside the cell was applied and the sample porosity measured. At each stress condition, we performed singlephase water flooding tests at several fixed rates (q_{w} = 1–5 mL/min) to obtain stressdependent absolute permeability. Afterward, the Hassler relative permeability method^{51,52}, in which water and N_{2} are simultaneously injected directly into the top of the core through two separate lines, was used to evaluate the steadystate pressure drop across the sample for each phase. These measurements were made at several water fractional flow values (f_{w} = 1, 0.8, 0.6, 0.4, 0.2, 0). For the capillary pressure measurement, we employed the modified stationary liquid method^{53}, keeping the wetting phase (water) immobile in the core while imposing several steadystate gas flow rates (q_{g} = 1–5 and 8 mL/min). Subtracting the core outlet pressure at steadystate condition from the gas pressure at the inlet provided an estimate of applied dynamic capillary pressure^{54}.
PoreScale Observations and Conceptual Proxy Modelling
Before running the corescale experiments, the carbonate specimen was scanned using a stationary Xray µCT at atmospheric pressure condition. The resulting 3D porescale model was synthetically compacted to achieve a proxy structure for the compacted core that was compatible with the imposed effective confining stress conditions. Figure 1 shows the work flow of the process implemented here. More detailed explanations are provided in the Materials and Methods.
Analytical Modelling of StressDependent Multiphase Flow Properties
As noted in^{46}, natural porous media are fractals, a fact which has led many researchers to adapt fractal scaling laws to describe the transport and other porerelated properties. Here, the fractal nature of pore sizes was used to develop analytical models for stressdependent capillary pressure and relative permeability curves, allowing for a better understanding of flow mechanisms on the scale of pores. Based on the fractal scaling laws, the cumulative sizedistribution (N(\(\bar{r}\))) of pores with diameter (\(\bar{r}\)) greater than or equal to r was assumed to follow the fractal scaling law given bellow^{55}:
where D_{f} and r_{max} are the fractal dimension and maximum pore diameter in porous media, respectively. Based on this definition, the equation for stressdependent capillary pressure P_{c}(σ′) can be shown to be
The terms P_{e}, ε_{p}, λ = 3 − D_{T} − D_{f} and D_{T} in eq. (2) refer to entry capillary pressure, pore strain, stressdependent pore size distribution index and tortuosity fractal dimension, respectively, and the index i indicates the initial stress condition (σ′ = 0) in this study. Based on the nonlinear stressstrain equation for the poroelastic materials, ε_{p} is related to the changes in effective confining stress (σ′) using the following equation^{56}, assuming pore strain (ε_{p}) to be equal to the volumetric strain (ε_{v}).
where A, B, and C are fitting constants. Haghi et al.^{34} defined the fitting constants in eq. (3) as A = (1 − γ_{s})/K_{H}, B = γ_{s}, and C = K_{S}, where K_{H} and K_{S} are the bulk moduli of hard and soft parts of the rock, respectively. γ_{s} refers to a ratio of the soft part volume to the bulk volume at an unstressed condition. Mixing eq. (2) with Burdine’s empirical equation^{19}, the following semianalytical equations are developed for relative permeability curves k_{r}(σ′).
where, S_{wir} and k_{rg−max} are irreducible water saturation and maximum gas relative permeability, respectively. A detailed derivation of the above equations is given in SI Appendix A.
Results and Discussions
CoreScale Approaches
The main contributions of the current research at the corescale are the experimental results illustrated in Fig. 2 of the stressdependent structural, singlephase flow, and multiphase flow properties of a porous carbonate rock under a wide range of isotropic effective confining stress conditions. In Fig. 2, circles and squares represent experimental data and colors indicate various effective confining stress conditions. Figure 2(a) shows the decline in stressdependent permeability (k) versus stressdependent porosity (φ) in response to an increasing effective stress condition, where the colored curve is fitted on the experimental data using the following equation based on the Carman and Kozeny correlation^{57}.
Here, τ defines the flow path hydraulic tortuosity inside the porous medium. The black circles in Fig. 2(b) represent the calculated normalized tortuosity (\(\bar{\tau }=\tau /{\tau }_{i}\)) from eq. (6) (at known stressdependent φ and k points) and the experimental pore strain data of the sample. The fitted curve is plotted based on the following equation^{34},
where the grain shape factor (a) and material constant (m) are calculated using the least squares regression technique (Fig. 2(b)). This plot clearly reveals an increasing intensity of flow path tortuosity under increasing effective stress conditions. This confirms the simultaneous dualeffect of stressinduced deformation on both pore and connectingchannels sizes, where these are comprehensible through pore strain and hydraulic tortuosity variation, respectively.
Fig. 2(c,d) illustrate the corescale experimental results of how the stressdependent multiphase flow properties vary under a range of effective confining stress condition from 10 MPa to 30 MPa. These two plots are an experimental demonstration of the remarkable controls that geomechanical processes impose on the multiphase flow properties of a porous carbonate rock. Herein, eq. (2) and (4–5) are applied as the curve fitting and interpolating equations. More details are given in SI Appendix A.
The results illustrated in Fig. 2(c,d) show leftward (decreasing S_{wir}) and upward shifts in the relative permeability and capillary pressure curves, respectively. These shifts are conceptually an indication of a decrease in pore sizes, which will be investigated in the next section.
Intuitively, one would expect that a porous medium’s increasing affinity to the gas phase should be responsible for the leftward shift of the relative permeability curves and decrease in S_{wir}. However, in our experiments, the interfacial tension and fluid properties were constant. To provide quantitative insight into the increasing affinity of the porous media to the gas phase, we provide important evidence from the experimental data at each stress condition: (1) irreducible water saturation (Fig. 3(a)), (2) water saturation at the intersection point of water and gas relative permeability curves (S_{wm}, Fig. 3(a)), and (3) maximum gas relative permeability (Fig. 3(b)). These observations, together with our knowledge of the nonchanging surface and interfacial tension between the phases and solid during the experiments, reveal the remarkable influence of pore and/or channel deformation on our understanding of the wettability of porous media. Figure 3(c) illustrates an increase in dynamic capillary pressure endpoints as effective stress is increased at constant flow rate conditions. The classical capillary number, defined as the ratio of the viscous forces to the capillary forces (i.e. \(Ca={\mu }_{g}{Q}_{inj}/\gamma A\))^{58}, increases slightly as the core crosssectional area (A) decreases in response to an increase in effective stress. For capillary number calculations, gas viscosity (\({\mu }_{g}=0.184\,mP.s\)), injection rate (\({Q}_{inj}=8\,mL/min\)), and interfacial tension (\(\gamma =69.36\,mN/m\)) are taken as stressindependent parameters. Mechanistically, the stressinduced increase of Ca (3.19–3.41 × 10^{−8}) under increasing effective stress conditions (0–30 MPa) improves the likelihood of capillary desaturation of the water phase in the porous media, based on the capillary desaturation curve (CDC) for drainage process^{58}; this is consistent with our experimental observations, in which S_{wir} declines as effective stress increases (Fig. 3(a)). However, given the limited range of calculated Ca for our experiments, in the next section further investigations are warranted for a more representative definition of the capillary number at the porescale with more reasonable water capillary desaturation values.
PoreScale Approaches
In this section, porescale manifestations are evaluated to postulate physical mechanisms supporting the significant corescale stressdependent multiphase flow properties measured in our experiments. The proxy modeling procedure given in Fig. 1 is used to develop conceptual 3D models of deformation at the porescale.
A qualitative demonstration of stressinduced capillary desaturation of the water phase in a single pore of our carbonate sample is presented in Fig. 4. Here, different colors indicate different effective stress conditions; the color bar corresponding to each is shown on the right side. 3D images of the selected pore at 0 MPa (gray) and 30 MPa (red) effective stress conditions are compared in the same frame in Fig. 4(a). The section in the solid black rectangle is magnified in Fig. 4(b–e) to depict the contraction of a channel connecting the main body to a branch pore with effective stress evolution from 0 MPa to 30 MPa.
Increasing effective stress, a simultaneous increase in both maximum applied capillary pressure (Fig. 3(c)) and channel capillary pressure occurs as the capillary radius r decreases (Fig. 4(b–e)) following the YoungLaplace equation \({P}_{c}=2\gamma \,\cos (\alpha )/r\)^{58}. When the applied corescale capillary pressure (\({P}_{c}^{Applied}={({P}_{g}{P}_{w})}_{sscond.}\)) becomes equal to the channel capillary pressure at a critical Ca, the gas viscous pressure (ΔP_{g}) becomes sufficient to expel, countercurrently, the water phase inside the pore through the channel. This leads to capillary desaturation of the water phase inside the pore. Gas viscous pressure is at its highest level at maximum applied capillary pressure (\({P}_{cmax}^{Applied}\)) under each effective stress condition where water saturation decreases to its minimum value at S_{wir}. Scaling down the classical Ca definition for a porechannel in a porous media, the terms viscous forces and capillary forces can be replaced with viscous pressure and capillary pressure of the channel, respectively^{59}. At critical state, maximum gas viscous pressure (ΔP_{gmax}) at maximum flow rate and steadystate condition, where water phase is immobile, is assumed to be equal to \({P}_{cmax}^{Applied}\)^{34,54}.
Equation (8) provides a new definition for the critical capillary number by connecting microscopic and macroscopic terms in a porous media to distinguish the onset of capillary desaturation of water at \(C{a}_{c}=1\) during the constant flow gas drainage process. Using eq. (8), Fig. 4(f), which provides a multifunctional phase redistribution diagram under variable effective stress conditions, illustrates new insights for stressdependent critical capillary numbers associated with the porechannel radius of the carbonate specimen used in this study.
At each effective stress condition, the threshold radius of the medium’s channels, in which the capillary desaturation of water starts, can be obtained at \(C{a}_{c}=1\). Additionally, the transition of a single channel, through its deformation path, from a gas capillary barrier to a gas flow conduit (or reverse) at different stress conditions can be recognized. With respect to the channel illustrated in Fig. 4(b–e), increasing effective stress from zero to 10 MPa (Fig. 4(b,c)) results in a transition in the state of the channel from a gas barrier to gas flow conduit; this leads to capillary desaturation of the water phase inside the branch pore. This microscopic transition explains the macroscopic decrease in S_{wm} and S_{wir} in response to an increase in effective stress (Fig. 3(a)). A decrease in S_{wm} and S_{wir} can be translated into an increasing affinity of porous media for the gas phase by increasing effective stress.
To provide further quantitative description of stressdependent changes in porescale structure, pore size distribution, with pore radius (r) being ≥4 μm (Fig. 5(a)), and two metrics, namely 1) 3D fractal dimension (D_{f}) which quantifies how 3D objects fills pore spaces (Fig. 5(b)) and 2) degree of anisotropy (D_{A}), which is a measure of 3D symmetry of pores in the media (Fig. 5(c)), are calculated at each effective stress condition. Details on these calculations are given in the Materials and Methods.
As a result of increasing the effective confining stress from 0 MPa to 30 MPa, a leftward shift in effective pore size distribution (PSD) is observed (Fig. 5(a)), while the mean pore radius decreases from 70 μm to 67 μm, respectively. Additionally, the cumulative frequency of pores with r ≤ 52 μm increases from 49% at 0 MPa effective stress to 52% at 30 MPa effective stress; the remaining pores span up to 326.85 μm and 275.25 μm, respectively. These stressdependent declines in the mean pore radius and size of the largest effective pore, pore connected to the surface of the core, result in an increasing trend of capillary pressure curves and entry capillary pressure, respectively^{58}, in response to an increase in effective stress. A similar trend can be realized at the corescale using the stressdependent capillary pressure experimental results shown in Fig. 2(d). Additionally, the leftward shift in PSD justifies the decline in maximum gas relative permeability under increasing effective stress conditions (Fig. 2(b)).
A declining trend was observed in both dimensionless numbers D_{f} and D_{A} in response to an increase in effective stress from 0 MPa to 30 MPa (Fig. 5(b,c)). These two pore structural metrics quantify clearly the rock deformation and its impacts on singlephase and multiphase parameters. The decreasing trend of D_{f} is an indication of more uniform (i.e. wellsorted) pore sizes^{58}; its striking impact on decreasing porosity, absolute permeability, and S_{wir} and increasing slope of capillary pressure curves has been shown experimentally and analytically in the literature^{24,46,47,48,49,50}. In a similar way, our calculated D_{A} represents an increasing state of isotropy of the sample due to pore compaction in response to increasing effective stress.
Implicational Insights
Simultaneous flow of gases and liquids in natural porous media is an indispensable phenomenon to study when we are dealing with subsurface flow processes. Stressdependent shifts in the relative permeability and capillary pressure curves have a major impact on our classical methods of understanding, quantifying, and modeling multiphase fluid flow in subsurface formations for a wide range of engineering and natural applications, including enhanced oil recovery, CO_{2} subsurface storage, heat energy recovery from geothermal reservoirs, and surface water infiltration in the vadose zone.
Although in this study we have conducted the flooding experiments on a standard carbonate specimen using water (wetting phase) and N_{2} (nonwetting phase) phases, the discovered systematic stressdependent behavior of rock could be extended for a wider range of rocks such as sandstone^{34} and fluid pairs (e.g. gas + liquid), as gases are typically the nonwetting phase for most natural porous media. For instance, the stressdependent upward shift in the entry capillary pressure for the studied carbonate rock, at an increasing effective stress condition from 10 MPa to 30 MPa (Fig. 3(c)), almost doubles the injection pressure and energy required to push the gas phase into the porous rock. However, the stressdependent decrease in irreducible wetting phase saturation resulting from increasing the effective stress condition from 10 MPa to 30 MPa (Fig. 3(a)) leads to an almost 20% increase in the volume of stored gas inside the core and recovery of the wetting phase (i.e. water) from the core. These dramatic changes in gas injection pressure, volume of stored gas, and recovery of the wetting phase are technically and economically significant for similar scenarios such as CO_{2} subsurface storage and enhanced recovery of hydrocarbon reservoirs using an miscible/immiscible gas (e.g. CO_{2}) flooding technique. These stressdependent processes allow us to predict additional CO_{2} storage in subsurface formations. In the same way, stressdependent leftward shifts in the relative permeability curves facilitate rainwater infiltration into the vadose zone and hot water flow through twophase geothermal reservoirs. This fundamental study paves the way for a more realistic study of fluid flow mechanisms in any natural and industrial applications which are dealing with multiphase fluid flow properties of rock under variable effective stress conditions.
Summary and Conclusions
We have shown experimentally the systematic impact of pore deformation as a result of effective stress changes on the singlephase and multiphase fluid flow properties of a carbonate specimen via a series of coreflooding experiments using water/N_{2} phases under isothermal and triaxial isotropic effective stress conditions. Our experiments have revealed leftward and upward shifts in the relative permeability and capillary pressure curves, respectively, with increasing effective stress, and have shown a decreasing shift of S_{wir}, S_{wm}, and k_{rwmax} under increasing effective stress conditions. These experiments have provided us corescale insights into the linkage between these shifts to decreases in porosity and absolute permeability and increases in pore strain and pore flow tortuosity under an increasing effective stress condition. Fractal and poroelasticity theories were used to drive analytical equations for stressdependent relative permeability and capillary pressure curves. These equations were used for curve fitting and interpolation in this study.
Using Xray computed microtomography and a proxy modelling technique, we were able to quantify the structure and shape of the pores and channels, pore size distribution, and scaling dimensionless numbers (e.g. fractal dimension and degree of anisotropy) at different effective stress conditions. We have shown that increasing effective stress leads to a leftward shift in PSD and closure of microchannels. As revealed in the 3D phase redistribution diagram (Fig. 4(f)), we have quantified the stressdependent threshold radius of the medium’s channels at the starting point of capillary desaturation of water in the porous media. Additionally, we have revealed the transition of a single channel, through its deformation path, from a gas capillary barrier into a gas flow conduit in response to an increasing stress conditions, which resulted in the capillary desaturation of the water phase and a macroscopic decrease in S_{wm} and S_{wir}. We have further revealed that leftward shift in PSD is responsible for the increasing trend of capillary pressure curves under an increasing effective stress condition. D_{f} and D_{A} metrics proved quantitatively the role of effective stress condition on pore size distribution and pore isotropy, respectively, in porous media.
These findings underscore the significant impact of effective stress on multiphase flow properties of rock, and manifest the physical mechanisms, such as capillary desaturation and capillary gas blockage due to stressinduced mechanical deformation of pores and/or channels, that control those properties.
Materials and Methods
CoreScale Experiments
The experiments were conducted on a waterwet Indiana limestone specimen (from Kocurek Industries INC., USA) 3.81 cm in diameter and 10.16 cm in length. The uniaxial compressive strength and Young’s modulus of the rock were 25.78 MPa and ≈3 GPa, respectively (Fig. S1). N_{2} and deionized water were used as the flooding phases and CO_{2} as the displacing phase. Here, an airwater contact angle (α) of 18.8 degrees were measured using a drop shape analyzer (DSA), which indicated the strong waterwet nature of the specimen (Fig. S2). Full resaturation of the confined sample inside the cell was a key challenge that we tackled using a procedure consisting of 3 steps: 1) flushing the core with 50 pore volume of high pressure CO_{2}, 2) vacuuming the sample (at ambient water saturation pressure) from an outlet line over a liquid nitrogen cold trap until a plateau pressure was achieved at the inlet, and 3) injecting high pressure water into the sample and maintaining the pressure for a few hours before reflushing it with 50 pore volume of water. The experimental procedure in this study is analogous to the protocol described in ref.^{34} for Berea sandstone, to which readers are referred for additional details.
Imaging and Proxy Modelling
The Xray microtomography was performed using the microCT imaging suite in the Pharmaceutical Orthopedic Research Lab (PORL) at the University of Alberta. The Xray source within this equipment is a sealed tube with a voltage of 100 kV and a spot size of <5 µm. The imager scans 360 degrees rotational fieldofview at a time using a 12 bit Xray detector that is fiberoptically coupled to a scintillator. We scanned an 18 mm length of the fulldiameter core 5 mm away from the end face of the core. All tomographic images were at 8.6 μm per voxel resolution. The postprocessing and quantifications were performed using the CT Analyser program within Bruker microCT 3D Suite software. The proxy modeling process in this study consisted of 7 steps, and the pore stressstrain plot of the rock from the corescale experiments was an initial input in this cycle. In the first step, we chose an effective stress condition and identified the corresponding pore strain from our corescale experiments (\({\varepsilon }_{{p}_{EXP}}\)). Then the original Xray µCT images were filtered using median and contrast enhancement techniques. Filtered images were segmented into black and white binary areas representing pores and grains, respectively. Through the thresholding step, a grayscale index smaller than its original value for the rock at zero effective stress was selected manually. Otsu’s thresholding method^{60} in 3D space was used initially to define the original grayscale index and match the model’s 3D porosity with experimentally measured porosity at zero effective stress. The model’s pore system was then assigned as a 3D object through bitwise operation, and its volume and strain (\({\varepsilon }_{{p}_{M}}=({V}_{{p}_{Mi}}{V}_{{p}_{M}})/{V}_{{p}_{Mi}}\)) were calculated. \({V}_{{P}_{Mi}}\) and \({V}_{{P}_{M}}\) defined as proxy model’s initial and stressdependent pore volume, respectively. This process was repeated iteratively until the pore strain of the proxy model became equal (±0.1%) to its corresponding experimental value and the representative model used for further 3D analysis.
Pore Size Distribution and Pore Metrics
We implemented the 3D calculation of pore size distribution, fractal dimension, and degree of anisotropy using the CT Analyser program within Bruker microCT 3D Suite Software. For the pore network process, firstly, a “skeletonisation” was performed to identify the medial axis of all pores. Secondly, a “spherefitting” measurement was made for all of the voxels along this axis. Pore size is defined as the diameter of the largest sphere which meets two conditions: (1) the sphere encloses the point (but the point is not necessarily the center of the sphere), and (2) the sphere is entirely located within the pore. The Kolmogorov or boxcounting method was used to calculate D_{f} in 3D; following this process, the volume was divided into an array of equal cubes and the number of cubes containing part of the object was counted. The process was repeated over a range of box sizes (3 to 100 pixels). Plotting the box sizes (L) versus the number of occupied boxes (N(L)), the slope of loglog linear regression defined D_{f}. Note that a linear relationship between the number of occupied boxes and box sizes in a loglog plot with the slope of D_{f} is also a confirmation of fractal nature of the porous sample based on \(N(L)\propto {L}^{{D}_{f}}\), where this was the case in the current study with the slopes given in Fig. 5(b). Finally, mean intercept length (MIL) analysis was used to measure D_{A}. As part of this process, a grid of lines was sent through the 3D core image containing pores (object) and grains (void) over a large number of 3D angles, and the length of the test line through the analyzed volume was divided by the number of times that the line passed through or intercepted pores. For more details, readers are referred to the software manual (https://www.bruker.com/products/microtomography.html).
References
Bear, J. Dynamics of Fluids in Porous Media. (Elsevier, New York) (1972).
Adler, P. M. & Brenner, H. Multiphase flow in porous media. Annual review of fluid mechanics 20(1), 35–59 (1988).
Bodvarsson, G. S., O’Sullivan, M. J. & Tsang, C. F. The sensitivity of geothermal reservoir behavior to relative permeability parameters (No. SGPTR5031). Earth Sciences Division, LBNL (Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA) (1980).
Gudjonsdottir, M., Palsson, Eliasson, H. J. & Saevarsdottir, G. Calculation of relative permeabilities from field data and comparison to laboratory measurements. Geothermics 54, 1–9 (2015).
CuetoFelgueroso, L. & Juanes, R. Nonlocal interface dynamics and pattern formation in gravitydriven unsaturated flow through porous media. Physical Review Letter 101(24), 244504 (2008).
Szulczewski, M. L., MacMinn, C. W., Herzog, H. J. & Juanes, R. Lifetime of carbon capture and storage as a climatechange mitigation technology. Proceeding of the National Academy of Science 109(14), 5185–5189 (2012).
Bachu, S. Review of CO2 storage efficiency in deep saline aquifers. International Journal of Greenhouse Gas Control 40, 188–202 (2015).
Honarpour, M. & Mahmood, S. M. Relativepermeability measurements: An overview. Journal of petroleum technology 40(08), 963–966 (1988).
Oak, M. J., Baker, L. E. & Thomas, D. C. Threephase relative permeability of Berea sandstone. Journal of Petroleum Technology 42(08), 1–054 (1990).
Rabbani, H. S., JoekarNiasar, V., Pak, T. & Shokri, N. New insights on the complex dynamics of twophase flow in porous media under intermediatewet conditions. Scientific reports 7(1), 4584 (2017).
Akbarabadi, M. & Piri, M. Relative permeability hysteresis and capillary trapping characteristics of supercritical CO2/brine systems: An experimental study at reservoir conditions. Advances in Water Resources 52, 190–206 (2013).
Singh, K. et al. Dynamics of snapoff and porefilling events during twophase fluid flow in permeable media. Scientific reports 7(1), 5192 (2017).
Datta, S. S., Dupin, J. B. & Weitz, D. A. Fluid breakup during simultaneous twophase flow through a threedimensional porous medium. Physics of Fluids 26(6), 062004 (2014).
Pak, T., Archilha, N. L., Mantovani, I. F., Moreira, A. C. & Butler, I. B. The dynamics of nanoparticleenhanced fluid displacement in porous mediaa porescale study. Scientific reports 8(1), 11148 (2018).
Rabbani, H. S., Zhao, B., Juanes, R. & Shokri, N. Pore geometry control of apparent wetting in porous media. Scientific reports 8(1), 15729 (2018).
Armstrong, R. T., Porter, M. L. & Wildenschild, D. Linking porescale interfacial curvature to columnscale capillary pressure. Advances in Water Resources 46(0), 55–62 (2012).
Raeini, A. Q., Blunt, M. J. & Bijeljic, B. Direct simulations of twophase flow on microCT images of porous media and upscaling of porescale forces. Advances in Water Resources 74(0), 116–126 (2014).
Purcell, W. R. Capillary pressurestheir measurement using mercury and the calculation of permeability therefrom. Journal of Petroleum Technology 1(02), 39–48 (1949).
Burdine, N. Relative permeability calculations from pore size distribution data. Journal of Petroleum Technology 5(03), 71–78 (1953).
Brooks, R. H. & Corey, A. T. Properties of porous media affecting fluid flow. Journal of the Irrigation and Drainage Division 92(2), 61–90 (1966).
Van Genuchten, M. T. A closedform equation for predicting the hydraulic conductivity of unsaturated soils 1. Soil science society of America journal 44(5), 892–898 (1980).
Pan, X., Wong, R. C. & Maini, B. B. Steadystate twophase in a smooth parallel fracture, paper 9639. The Petroleum Society in Calgary, Alberta (1996).
Yang, Z., Niemi, A., Fagerlund, F. & Illangasekare, T. Twophase flow in roughwalled fractures: comparison of continuum and invasionpercolation models. Water Resources Research 49, 993–1002, https://doi.org/10.1002/wrcr.20111 (2013).
Li, K. & Horne, R. N. Universal capillary pressure and relative permeability model from fractal characterization of rock. Stanford Geothermal Workshop on Geothermal Reservoir Engineering, Stanford University, Stanford, California (2004).
Montessori, A. et al. Effects of Knudsen diffusivity on the effective reactivity of nanoporous catalyst media. Journal of Computational Science 17, 377–383 (2016).
Falcucci, G. et al. Mapping reactive flow patterns in monolithic nanoporous catalysts. Microfluidics and Nanofluidics 20(7), 105 (2016).
Montessori, A., Prestininzi, P., La Rocca, M. & Succi, S. Lattice Boltzmann approach for complex nonequilibrium flows. Physical Review E 92(4), 043308 (2015).
Montemore, M. M. et al. Effect of nanoscale flows on the surface structure of nanoporous catalysts. The Journal of chemical physics 146(21), 214703 (2017).
McDonald, A. E., Beckner, B. L., Chan, H. M., Jones, T. A. & Wooten, S. O. Some important considerations in the simulation of naturally fractured reservoirs. In Low Permeability Reservoirs Symposium. Society of Petroleum Engineers (1991).
Khan, A. H. Shear Induced Relative Permeability Changes in Uncemented Sands. Masters of Science in Engineering Thesis, December 2009, University of Texas at Austin, USA (2009).
Hamoud, M., Chalaturnyk, R. J. & Leung., J. Influence of Geomechanical Processes on Relative Permeability. Canadian Rock Mechanics Symposium, Edmonton, Alberta, 9 p (2012).
Lian, P., Cheng, L. & Ma, C. Y. The characteristics of relative permeability curves in naturally fractured carbonate reservoirs. Journal of Canadian Petroleum Technology 51(02), 137–142 (2012).
Huo, D. & Benson, S. M. Experimental Investigation of StressDependency of Relative Permeability in Rock Fractures. Transport in Porous Media, 1–24 (2016).
Haghi, A. H., Talman, S. & Chalaturnyk, R. Consecutive Experimental Determination of Stressdependent Static and Dynamic Properties of Berea Sandstone and Implications for Twophase Flow Modelling. Water Resources Research (2019).
Ofagbohunmi, S., Chalaturnyk, R. J. & Leung, J. Y. W. Coupling of StressDependent Relative Permeability and Reservoir Simulation. In SPE Improved Oil Recovery Symposium. Society of Petroleum Engineers (2012).
Lei, G., Mo, S., Dong, Z., Wang, C. & Li, W. Theoretical and experimental study on stressdependency of oil–water relative permeability in fractal porous media. Fractals 26(02), 1840010 (2018).
Haghi, A. H., Chalaturnyk, R. & Geiger, S. New Semi‐Analytical Insights Into Stress‐Dependent Spontaneous Imbibition and Oil Recovery in Naturally Fractured Carbonate Reservoirs. Water Resources Research 54(11), 9605–9622 (2018).
Haghi, A. H., Chalaturnyk, R. & Ghobadi, H. The state of stress in SW Iran and implications for hydraulic fracturing of a naturally fractured carbonate reservoir. International Journal of Rock Mechanics and Mining Sciences 105, 28–43 (2018).
Haghi, A. H., Kharrat, R., Asef, M. R. & Rezazadegan, H. Presentday stress of the central Persian Gulf: implications for drilling and well performance. Tectonophysics 608, 1429–1441 (2013).
Haghi, A. H., Kharrat, R. & Asef, M. R. A case study for HCLbased fracturing and stress determination: A Deformation/Diffusion/Thermal approach. Journal of Petroleum Science and Engineering 112, 105–116 (2013).
Detournay, E. & Cheng, A. H. D. Fundamentals of poroelasticity, comprehensive rock engineering: principles, practice and projects. Analysis and design method (1993).
Biot, M. A. General theory of three‐dimensional consolidation. Journal of applied physics 12(2), 155–164 (1941).
Jones, C., AlQuraishi, A., Somerville1, J. & Hamilton, S. Stress Sensitivity of Saturation and EndPoint Relative Permeability. Society of Core Analysts (2001).
Li, J. J., Liu, Y., Gao, Y. J., Cheng, B. Y. & Jiang, H. Q. Porescale study of the pressuresensitive effect of sandstone and its influence on multiphase flows. Petroleum Science, 1–14 (2019).
Li, P., Chalaturnyk, R. J. & Polikar, M. Issues with reservoir geomechanical simulations of the SAGD process. Journal of Canadian Petroleum Technology, 43(05) (2004).
Yu, B., Li, J., Li, Z. & Zou, M. Permeabilities of unsaturated fractal porous media. International journal of multiphase flow 29(10), 1625–1642 (2003).
Zheng, X., Zhigang, C. & Weichuan, L. Gas/water flowing ability influence experimental study of permeability stress sensibility in tight gas reservoir. Well Logging Technology 4, 360–363 (2013).
Xu, P., Qiu, S., Yu, B. & Jiang, Z. Prediction of relative permeability in unsaturated porous media with a fractal approach. International Journal of Heat and Mass Transfer 64, 829–837 (2013).
Mo, S. Y., He, S. L., Lei, G., Gai, S. H. & Liu, Z. K. Effect of the drawdown pressure on the relative permeability in tight gas: A theoretical and experimental study. Journal of Natural Gas Science and Engineering 24, 264–271 (2015).
Gao, H., Yu, B., Duan, Y. & Fang, Q. Fractal analysis of dimensionless capillary pressure function. International Journal of Heat and Mass Transfer 69, 26–33 (2014).
Osoba, J. S., Richardson, J. G., Kerver, J. K., Hafford, J. A. & Blair, P. M. Laboratory measurements of relative permeability. Journal of Petroleum Technology 3(02), 47–56 (1951).
Richardson, J. G., Kerver, J. K., Hafford, J. A. & Osoba, J. S. Laboratory determination of relative permeability. Journal of Petroleum Technology 4(08), 187–196 (1952).
Leas, W. J., Jenks, J. H. & Russell, C. D. Relative permeability to gas. Journal of Petroleum Technology 2(03), 65–72 (1950).
Pini, R. & Benson, S. M. Simultaneous determination of capillary pressure and relative permeability curves from core‐flooding experiments with various fluid pairs. Water Resources Research, 49(6) (2013).
Yu, B. & Cheng, P. A fractal permeability model for bidispersed porous media. International Journal of Heat and Mass Transfer 45(14), 2983–2993 (2002).
Zimmerman, R. W., Somerton, W. H. & King, M. S. Compressibility of porous rocks. Journal of Geophysical Research 91(B12), pp.12765–12777 (1986).
Carman, P. C. Flow of gases through porous media. Academic press (1956).
Peters, E. J. Advanced Petrophysics: Dispersion, interfacial phenomena (Vol. 2). Greenleaf Book Group (2012).
Hilfer, R. & Øren, P. E. Dimensional analysis of pore scale and field scale immiscible displacement. Transport in Porous Media 22(1), 53–72 (1996).
Otsu, N. A threshold selection method from graylevel histograms. IEEE transactions on systems, man, and cybernetics 9(1), 62–66 (1979).
Acknowledgements
We thank Kevin Hodder from GeoPrint RG^{2} group for assisting us with DSA equipment access and droplet imaging, Alex Peeters from GeoREF RG^{2} group for helping us with the initial version of lab equipment design, and Micheal Doschak from OMX Imaging Services for his guidance on µCT image processing. The authors also acknowledge the financial support of the Energi Simulation Industrial Research Consortia on Reservoir Geomechanics.
Author information
Authors and Affiliations
Contributions
A.H. designed the research, conducted the experiments, analyzed the data, and wrote the first manuscript. R.C. helped in developing the idea of the study and revised the manuscript. He supervised the general aspects and progression of the project. A.H. is a PhD student in R.C.’s research group. S.T. contributed to experimental design and revised the manuscript. All authors reviewed the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Haghi*, A.H., Chalaturnyk, R. & Talman, S. StressDependent Pore Deformation Effects on Multiphase Flow Properties of Porous Media. Sci Rep 9, 15004 (2019). https://doi.org/10.1038/s41598019512630
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598019512630
 Springer Nature Limited
This article is cited by

Radon transport carried by geogas: prediction model
Environmental Science and Pollution Research (2023)

PoroMechanical Coupling for Flow Diagnostics
Transport in Porous Media (2022)