Reconciling the Long‐Term Relationship Between Reservoir Pore Pressure Depletion and Compaction in the Groningen Region

Abstract

The Groningen gas reservoir, situated in the northeast of the Netherlands is western Europe's largest gas reservoir. Due to gas production measurable subsidence and seismicity has been detected across this region, attributed to the deformations induced by reservoir pore pressure depletion. We investigate the surface displacement history using a principal component analysis‐based inversion method to combine a diverse set of optical leveling, interferometric synthetic aperture radar, and Global Positioning System data to better constrain reservoir compaction and subsidence history. The generated compaction model is then used in combination with prior pressure depletion models to determine a reservoir uniaxial compressibility. The best fitting model of uniaxial compressibility is time independent but spatially variable. The absence of evidence for any significant time delay between changes in depletion and compaction rates supports an instantaneous poroelastic reservoir response. The absence of nonlinear yielding at the largest compaction strains suggests that anelastic deformations are a minor part of reservoir compaction.

Time-Dependent Stresses From Fluid Extraction and Diffusion With Applications to Induced Seismicity

“…Since seismicity can also often be observed in the layers overlying and surrounding the producing or injection zone, the focus in this study is to expand the analyses of Segall [30,32] to examine the evolution of the stress field directly surrounding the producing zone in a reservoir including the smooth temporal evolution of the pore fluid distribution through diffusion. While the fixed reservoir model of Segall [30] was shown to qualitatively agree well with patterns in surface deformation and seismicity surrounding reservoirs with potentially comparable reservoir geometries, we note that the particular example of a laterally unconstrained reservoir with a finite diffusivity may be more quantitatively applicable to other producing fields where changes in fluid mass may not be considered uniform within the producing region (e.g., central Oklahoma [15], North and South Dakota [33], and Groningen [34]). Moreover, for many reservoir settings, the timescale for fluid transport (days to years) within the producing layer may be comparable to the timescale of field observations at relevant distances.…”

Stress‐Dependent Magnitudes of Induced Earthquakes in the Groningen Gas Field

“…The activity rate model parameters β 0 , β 1 describe an exponential relationship between Poisson intensity and incremental stress. the training data and is commensurate with the reservoir depth that governs the lateral resolution of reservoir strains inferred from the observed surface displacements [Smith et al, 2019]. The optimal fault filtering, r max = 0.4, corresponds to excluding faults with offsets that reduce reservoir juxtaposition by 40%.…”

“…This has steadily reduced the initial mean pore pressure by up to 25 MPa within the gas-bearing sediments of the Upper Rotliegend (Permian) and Limburg (Carboniferous) Groups [Stauble and Milius, 1970] that are located 2600-3200 m below the surface. The resulting reservoir compaction has steadily increasing surface subsidence up to 400 mm as observed by optical leveling surveys since 1964 [Bourne et al, 2014;Smith et al, 2019] and InSAR since 1995 [Ketelaar, 2008[Ketelaar, , 2009. These reservoir deformations reactivate intra-reservoir faults that have been inducing observable seismicity since at least 1986 [Dost et al, 2012] with an exponential-like increase in the cumulative number of earthquakes relative to the cumulative volume of gas produced [Bourne and Oates, 2017b;Bourne et al, 2018].…”

Microphysics of Inelastic Deformation in Reservoir Sandstones from the Seismogenic Center of the Groningen Gas Field

“…Even though these strains are small, they may induce sufficiently high shear tractions on pre-existing faults to trigger seismogenic rupture (Mulders 2003;Bourne et al 2014;Buijze et al 2017). Constraining the magnitude of the in situ reservoir strain and particularly of the inelastic contribution is crucial for realistic geomechanical modeling of induced subsidence and seismicity Buijze et al 2017;Van Wees et al 2018;Candela et al 2019;Smith et al 2019). However, any inelastic sandstone deformation at the small strains (ε < 1%) relevant to induced reservoir compaction is frequently overlooked, since previous research (Wong and Baud 2012-and references therein) mainly focused on the deformation behavior seen at (much) higher strains (ε of 1-10%).…”