This paper presents an experimental study on the ability of organic-rich-shale core samples to store carbon dioxide (CO 2). An apparatus has been built for precise measurements of gas pressure and volumes at constant temperature. A new analytical methodology is developed allowing interpretation of the pressure/volume data in terms of measurements of total porosity and Langmuir parameters of core plugs. The method considers pore-volume compressibility and sorption effects and allows small gas-leakage adjustments at high pressures. Total gas-storage capacity for pure CO 2 is measured at supercritical conditions as a function of pore pressure under constant reservoir-confining pressure. It is shown that, although widely known as an impermeable sedimentary rock with low porosity, organic shale has the ability to store significant amount of gas permanently because of trapping of the gas in an adsorbed state within its finely dispersed organic matter (i.e., kerogen). The latter is a nanoporous material with mainly micropores (< 2 nm) and mesopores (2-50 nm). Storage in organic-rich shale has added advantages because the organic matter acts as a molecular sieve, allowing CO 2-with linear molecular geometry-to reside in small pores that the other naturally occurring gases cannot access. In addition, the molecular-interaction energy between the organics and CO 2 molecules is different, which leads to enhanced adsorption of CO 2. Hence, affinity of shale to CO 2 is partly because of steric and thermodynamic effects similar to those of coals that are being considered for enhanced coalbed-methane recovery. Mass-transport paths and the mechanisms of gas uptake are unlike those of coals, however. Once at the fracture/matrix interface, the injected gas faces a geomechanically strong porous medium with a dual (organic/inorganic) pore system and, therefore, has choices of path for its flow and transport into the matrix: the gas molecules (1) dissolve into the organic material and diffuse through a nanopore network and (2) enter the inorganic material and flow through a network of irregularly shaped voids. Although gas could reach the organic pores deep in the shale formation following both paths, the application of the continua approximation requires that the gas-flow system be near or beyond the percolation threshold for a consistent theoretical framework. Here, using gas permeation experiments and history matching pressure-pulse decay, we show that a large portion of the injected gas reaches the organic pores through the inorganic matrix. This is consistent with scanning-electron-microscope (SEM) images that do not show connectivity of the organic material on scales larger than tens of microns. It indicates an in-series coupling of the dual continua in shale. The inorganic matrix permeability, therefore, is predicted to be less, typically on the order of 10 nd. More importantly, although transport in the inorganic matrix is viscous (Darcy) flow, transport in the organic pores is not due to flow but mainly to molecular transport...

On the basis of micro-and mesoscale investigations, a new mathematical formulation is introduced in detail to investigate multiscale gas-transport phenomena in organic-rich-shale core samples. The formulation includes dual-porosity continua, where shale permeability is associated with inorganic matrix with relatively large irregularly shaped pores and fractures, whereas molecular phenomena (diffusive transport and nonlinear sorption) are associated with the kerogen pores. Kerogen is considered a nanoporous organic material finely dispersed within the inorganic matrix. The formulation is used to model and history match gaspermeation measurements in the laboratory using shale core plugs under confining stress. The results indicate significance of molecular transport and strong transient effects caused by gas/solid interactions within the kerogen. In the second part of the paper, we present a novel multiscale perturbation approach to quantify the overall impact of local porosity fluctuations associated with a spatially nonuniform kerogen distribution on the adsorption and transport in shale gas reservoirs. Adopting weak-noise and meanfield approximation, the approach applies a stochastic upscaling technique to the mathematical formulation developed in the first part for the laboratory. It allows us to investigate local kerogenheterogeneity effects in spectral (Fourier-Laplace) domain and to obtain an upscaled "macroscopic" model, which consists of the local heterogeneity effects in the real time-space domain. The new upscaled formulation is compared numerically with the previous homogeneous case using finite-difference approximations to initial/boundary value problems simulating the matrix gas release. We show that macrotransport and macrokinetics effects of kerogen heterogeneity are nontrivial and affect cumulative gas recovery. The work is important and timely for development of newgeneration shale-gas reservoir-flow simulators, and it can be used in the laboratory for organic-rich gas-shale characterization.

No abstract

In coalbeds and shales, gas transport and storage are important for accurate prediction of production rates and for the consideration of subsurface greenhouse gas sequestration. They involve coupled fluid phenomena in porous medium including viscous flow, diffusive transport, and adsorption. Standard approach to describe gas-matrix interactions is deterministic and neglects the effects of local spatial heterogeneities in porosity and material content of the matrix. In this study, adopting weak-noise and mean-field approximations and using a statistical approach in spectral domain, matrix heterogeneity effects are investigated in the presence of non-equilibrium adsorption with random partition coefficient. It is found that the local heterogeneities can generate non-trivial transport and kinetic effects which retard gas release from the matrix and influence the ultimate gas recovery adversely. Macrotransport shows 1/ [1 + N Pe /(1 + N Pe )] dependence on the Péclet number, and persists at the diffusive ultra-low permeability limit. Macro-kinetics is directly related to Thiele modulus by the following expression: N T h /(1 + 2N Pe ). It leads to trapping of gas in the adsorbed phase during its release from the matrix, and to an adsorption threshold during the gas uptake by the matrix. Both effects are proportional to the initially available adsorbed gas amount and becomes more pronounced with the increasing variance of the porosity field. Consequently, a new upscaled deterministic gas mass balance is proposed for practical purposes. Numerical results are presented showing free and adsorbed gas distributions and fractional gas sorption curves for unipore coal matrix exhibiting Gaussian porosity distribution. This study is a unique approach for our further understanding of the coalbeds and gas shales, and it is important for the development of sound numerical gas production and sequestration models. Absolute coal permeability (cm 2 ) C Free gas concentration (mol/cc pore) C µ Adsorbed gas concentration (mol/cc solid) C µs Maximum adsorbed gas concentration (mol/cc solid) D Molecular diffusion coefficient (cm 2 /s) D Apparent diffusion coefficient (cm 2 /s) E Adsorbate adsorbent interaction energy (J/mol) g Average free gas concentration (mol/cc) K Partition (distribution) coefficient (fraction) k f Gas adsorption rate coefficient (1/s) k r Gas desorption rate coefficient (1/s) k r∞ Gas desorption rate constant at zero energy level (1/s) R Universal gas constant (J K −1 mol −1 ) r Pore half width (cm) tTime coordinate (s)

This paper presents an experimental study on the ability of Barnett shale core samples to store carbon dioxide. An apparatus has been built for psrecise measurements of gas pressure and volumes at constant temperature. A new analytical methodology is developed allowing interpretation of the pressure-volume data in terms of measurements in total porosity and Langmuir parameters of core plugs. The method considers pore volume compressibility and sorption effects and allows small gas leakage adjustments at high pressures. Total gas storage capacity for pure carbon dioxide is measured at supercritical conditions as a function of pore pressure under constant reservoir confining pressure. It is shown that, although widely-known as an impermeable sedimentary rock with low porosity, organic shale has the ability to store significant amounts of gas permanently due to trapping of the gas in adsorbed state within its finely-dispersed organic matter, i.e., kerogen. The latter is a nanoporous material with micropores (&lt; 2 nm) and mesopores (2-50 nm). Storage in organic shale has the added advantages because the organic matter acts as molecular sieve allowing carbon dioxide —with linear molecular geometry— to reside in small pores that the other naturally-occurring gases cannot access. In addition, the molecular interaction energy between the organics and carbon dioxide molecules is different which leads to its enhanced adsorption. Hence, affinity of shale to carbon dioxide is due to partly steric and thermodynamic effects similar to those of coals that are being considered for enhanced coalbed methane recovery. Mass transport paths and the mechanisms of gas uptake are unlike coals, however. Once at the fracture-matrix interface, the injected gas faces a geomechanically strong porous medium with dual (organic/inorganic) pore system, therefore, has choices of path for its flow and transport into the matrix: the gas molecules (i) dissolve into the organic material and diffuse through a nanopore-network, and (ii) enter the inorganic material and flow through a network of irregularly shaped voids. Although the gas could reach the organic pores deep in the shale formation following both paths, the application of the continua approximation to the percolation threshold is not known. Here, using gas permeation experiments and history-matching pressure pulse decay, we show that a large portion of the injected gas reaches the organic pores through the inorganic matrix. This is consistent with SEM images that do not show connectivity of the organic material on scales larger than tens of microns. It indicates an in-series coupling of the dual continua in shale. The inorganic matrix permeability is therefore predicted less, typically in the order of 10 nD. More importantly, transport in the organic pores is not due to flow but mainly pore and surface diffusion mechanisms.