Effects of mechanical properties of the underburden on induced seismicity along a basement fault during hydrogen storage in a depleted reservoir

Abstract

This study models the geomechanical deformation of a depleted gas field, wherein gaseous hydrogen is stored in a North Sea reservoir, and is cyclically injected and withdrawn. A fault is modeled within the underburden, and its slip is investigated during a three year storage period. Parametric simulations are conducted to study the influence of the underburden mechanical properties, such as Young's modulus, Poisson's ratio, and permeability on induced seismicity. The fault is predominantly in stick during the bulk of the injection, storage, and withdrawal periods, but minor fault slip ( $<$ 4 mm) occurs shortly after a change in operational regime. The Young's modulus of the underburden unit has the strongest control on fault slip. To reduce the seismic hazard, an underburden with low Young's modulus ( $<$ 15 GPa), high Poisson's ratio ( $>$ 0.25), low Biot coefficient, and low permeability ( $<1\times {10}^{-19}$ m²) is found to be most suitable for hydrogen storage.

Keywords: Energy engineering; Petrophysics; Structural geology.

Graphical abstract

Figure 1

Depleted gas field model (A) A solid and wireframe view of the model showing the four rock layers (overburden is cream, caprock is light gray, reservoir is orange, underburden is dark gray). (B) A perspective front-right solid and wireframe view of the fault (blue) contained with the reservoir unit and the injection well on the left of the domain (orange) and the withdrawal well on the right of the domain (red). The reservoir and underburden units are removed in order to observe these. (C) A top view of the rectangular fault surface.

Figure 2

Comparison of different domain and fault meshes (A) Wireframe views of the four versions of the model for each refinement case in Table 4. The dashed boxes denote the zones of increased refinement. (B) Zoomed views of the mesh at the center of the underburden unit. (C) Mesh of the fault surface for the five versions of the fault in the model for each refinement case in Table 5, for varying amounts of fracture elements $E_{frac}$.

Figure 3

Plots of average slip as a function of operation time for different fault refinements, domain refinements and time steps (A) five cases of different fault refinements represented by different numbers of fault elements, (B) four cases of different bulk rock/domain refinements represented by different numbers of nodes and (C) four cases of different time step. The full discretization data for each case in (A) and (B) is reported in Tables 4 and 5 respectively.

Figure 4

Average slip as a function of operation time for four cases of varying underburden properties (A) Young’s modulus, (B) Poisson’s ratio, (C) Biot coefficient, and (D) permeability.

Figure 5

The apertures of the fault at a time of 1.58 years during the end of the second storage and beginning of the second withdrawal periods for cases of varying underburden properties (A–D) Young’s modulus, (E–H) Poisson’s ratio, (I–L) Biot coefficient, and (M–P) permeability.

Figure 6

The individual slip at each node of the fault at a time of 1.58 years during the end of the second storage and beginning of the second withdrawal periods for cases of varying underburden properties (A–D) Young’s modulus, (E–H) Poisson’s ratio, (I–L) Biot coefficient and (M–P) permeability. Note that the scales are not the same in each subfigure.

Figure 7

Plots of effective stress vertically through the center of the model in the z-direction for four cases of varying underburden properties (A) Young’s modulus, (B) Poisson’s ratio, (C) Biot coefficient, and (D) permeability at a time of 1.58 years.

Figure 8

The strain over a vertical interval containing the anticlinal caprock, reservoir and underburden units at a time of 1.58 years during the end of the second storage and beginning of the second withdrawal periods for cases of varying reservoir properties (A–D) Young’s modulus, (E–H) Poisson’s ratio, (I–L) Biot coefficient and (M–P) permeability.

Figure 9

Maximum potential induced magnitude versus average underburden fault slip Includes all potential events from the Young’s modulus, Poisson’s ratio, Biot coefficient, and permeability cases. Zoomed views of the two main event clusters from the left, are shown on the right.