8.3 Restoration, geomechanics, natural and induced fractures
As the industry continues migrating to more and more complicated plays, geomechanics is taking a more important place. Engineers might need to understand how to optimize the induced fractures they are creating along their horizontal wells. They might also need to understand how natural fractures is going to impact the production, either positively by improving the oil movement or negatively by moving deep water vertically along the faults and fractures. Geomechanics might also be central to engineering study on the layers above the reservoir and how the overburden might break because of the operations in the reservoirs.
All these issues require a characterization of the geomechanical properties of the reservoir. These properties can be modeled in the geomodel as well as the impact of natural fractures on the reservoir porosity and permeability.
The first set of techniques related to geomechanics are the extension to geomodels to the restoration techniques initially presented by (Dahlstrom, 1969). Dahlstrom’s key idea is that folding and faulting processes can be “reversed” to restore the rocks to their geometry before these deformations. It is also assumed that the layers were horizontal before deformation (at least for deformation that didn’t occur during deposition). Figure 8 shows the deformation of horizontal sedimentary layers into a fault-bend fold. Let’s assume that Figure 8C is the current geological interpretation. Following Dahlstrom’s restoration rules, one can unfold and unfault the structure to go back in time to the reservoir before deformation (Figure 8A). Such restoration technique can help spotting non-geological geometries in the current model. Figure 9A shows a schematic faulted structure. It is assumed that the faulting occurred after deposition. Once restored (Figure 9B), it appears clearer that the green layer is thicker in the hanging wall than in the footwall. The thickness of the blue layer is also not consistent across the fault. As the layers were deposited before faulting, there should be no change of thickness because of the faulting. It means the current interpretation is incorrect. Most likely, the geometry of the horizon being the top of the green layer must be changed.
While Dahlstrom’s approach was based on geometrical restoration, the whole concept was extended to include kinematics and rheological constraints. And nowadays, many geomodeling packages have implemented a form or another of restoration, not only of surfaces, but of full 3D-grids and even if underlying seismic cubes (in case the 3D geophysical interpretation was done in the geomodeling package in a way that the software as “connected” the seismic cube and the 3D-grid built from the fault and horizon interpretation).
Faulting and folding induce stress in the rocks which endure the deformation. For example, Figure 8C shows with grey dashes the part of the volumes where stress occurred and as such, where it is most likely to find natural fractures. In the context of geomodeling, it means that, in some reservoirs, a direct correlation can be made between some part of the curvature of the model and the presence of natural fractures. This information can then be coupled with fracture characterization done on seismic data.
Modeling natural fractures require to combine data from multiple sources. Fracture characterization along well cores or well FMI allow to define the fracture density along the well (= number of fracture by unit of length along the well) as well as details about the size of fractures (length, aperture) and their orientation. Well data will also allow understanding if the reservoir contain one or several fracture sets. Each fracture set has its on parameters (in a similar way than different facies have specific ranges of porosity or permeability). Each fracture set might be correlated to a different stage of deformation in the life of the reservoir.
The fracture characterization done at the wells is then correlated with seismic attributes, especially to find a signature to the fracture density in the seismic. If such seismic attributes are identified, they can be used as a guide, a secondary variable, in the modeling of the fracture density of each set between the wells. In geomodeling terms, modeling fracture density imply similar processes than modeling porosity for example with geostatistical tools like sequential simulation with a seismic attribute as a secondary property. As such, the reader can refer to the other chapters of this book for a review of these methods.
Once the different fracture sets parameters are modeled in 3D, they are used to compute fracture porosity and permeability, which effects influence the over porosity and permeability of the reservoir. All this information about natural fractures are useful for flow simulation. They are also needed as input for fracture modeling done by production engineers. Their goal is to model how induced fractures are going to grow in the reservoir. These predictions are used to optimize fracking.