Effects of Anisotropic Poroelasticity on Stress and Pore Pressure Changes Around Subsurface Reservoirs and Storage Sites

Abstract

Stresses and pore pressures change around depleting hydrocarbon reservoirs or fluid injection sites as result of reservoir pore pressure change. Such changes influence the stability of new wells drilled and may lead to fault activation, resulting in induced seismicity and possibly to casing damage in existing wells. Forecasting requires reliable geomechanical modeling of the reservoir and its surroundings. Classical analytical solutions are too simple, so numerical models are required. The rocks surrounding reservoirs are often anisotropic (like overburden shale), but their anisotropic elasticity is normally neglected, either because of model limitations, or in lack of input data. Here, however, the impact of anisotropy along with the elastic contrast between undrained surroundings and drained (isotropic) reservoir rock is addressed through a series of numerical scenarios. It is shown that isotropic models are unable to reproduce the stress changes predicted by the anisotropic models. Pore pressure changes may be underestimated by neglecting the anisotropic nature of the poroelastic Skempton parameters.

Material Constants of Anisotropic Poroelasticity and Its Impacts on Shale Gas Production

“…While the bedding plane orientation has an evident effect on the horizontal stress parallel to the wellbore (in our case, σ xx ), its effect on σ yy is less significant, as shown by Figure 9. This finding is quite consistent with the findings from the work on rock anisotropy, 4,10,17,25 i.e., as opposed to the habitually applied isotropic assumption, the horizontal stress parallel to the wellbore shows a strong dependency on the anisotropic Figure 5. Three different mesh sizes in the mesh sensitivity analysis of the 3D problem.…”

“…Although in Sections and 3, we assume the most general condition, in the following numerical simulations, we still adopt the ideal porous medium assumption due to lack of reasonable complete input data, so K x = K ψ = ∞, and there will be no distinction among K s , K̅ s , K s ′, and K s ″. Hence, we simply denote them as K s (solid grain bulk modulus) for brevity.…”

“…However, shale, as one of the most widespread sedimentary formations on Earth, exposed well-defined anisotropy in the form of bedding, lamination, foliation (repetitive layering), fissuring, or jointing. , The fissures or joints could generate anisotropic and preferential fluid flow paths , and lead to a certain level of discontinuity, depending on the size of the problem of interest . For example, for a regional problem on the reservoir scale, the associated deformation discontinuity across enormous fissures and joints could be ignored, while they mostly contribute to the fluid flow through the so-called multiple porosity model. ,– Therefore, in this study, we focus on the so-called intact anisotropy, which means that for any material point in the domain, the stiffness and/or permeability vary with the observing direction. ,, This anisotropy can have a significant impact on the way that fractures propagate through the rock, , the direction in which gas flows, , and the shear failure. – Therefore, coupled hydro-mechanical simulations should take the anisotropic characteristics into account. ,– …”

Anisotropic Poroelastic Modelling of Depletion-Induced Pore Pressure Changes in Valhall Overburden

“…Despite these large percentage changes, the inclusion of the poroelastic anisotropy should not have much impact on pore pressure predictions in the part of the overburden further away the reservoir, as the corresponding absolute magnitude of the undrained pore pressure change given by both of the two predictions is small. In the case of isotropic geomechanical model of the overburden and isotropic poroelastic pore pressure parameters, the expected differences in predicted undrained pore pressure responses are significantly larger (Holt et al 2022). Next, we use Eq.…”