**7.1 2D effect of weak interface on hydraulic fracture height**

A specimen, under high vertical overburden stress, made of three layers alternating soft and stiff rock as estimated from the surface drilling estimation of Poisson's ratio and Young's Modulus is used to highlight the potential of a hydraulic fracture to develop a step-over behavior. We consider two cases: case 1 has perfect interfaces and case 2 has weak interfaces. To illustrate the step-over phenomenon, the two cases include a flaw (small fracture) located at the interface, close to the vertical path from the injection point (**Figure 11**).

 **Figure 11** shows that for case 1 with a strong interface, the vertical fracture growth was insensitive to the presence of the flaw and the fracture propagated in the direction of the applied stress giving rise to symmetric fracture half-heights (**Figure 11**, top). However, for case 2 with the weak interface and a flaw near the injection point, the hydraulic fracture was first arrested by the weak interfaces, and then stepping over occurred when the fracture reached the interface (**Figure 11**, bottom). The fracture height is much smaller in the case 2 with weak interfaces and a strong asymmetric fracture height was developed as a result of high shear in the weaker interfaces.

The weak interfaces gave rise to an asymmetric fracture half-height, where the flaw promoted the propagation toward the direction of its location. These flaws, which are mainly bed-bounded natural fractures, could affect the propagation path of the hydraulic fracture and generate asymmetric hydraulic fracture half-heights or arrest the parent fracture and promote the secondary child fracture. It is this complex geologic reality that makes current hydraulic fracturing simulators inadequate to capture the propped frac height. This geologic complexity gets more complicated as we consider actual 3D situations.

### **7.2 3D effects of dipping fractures planes on fracture geometry**

 There are multiple field conditions that cannot be modeled with 2D approximations, and full 3D modeling is needed. This is the case in a strike-slip stress regime, when dealing with natural fracture planes that are not vertical, or when stimulating with helical perforation, etc. In these cases, full 3D modeling tools are essential to accurately reproduce the 3D fracture propagation mechanisms that will lead to the

 correct fracture height. To illustrate a full 3D analysis, a 3D laboratory test in [27] was considered. Briefly, the laboratory specimen was made with a real reservoir sandstone. The specimen, subjected to the strike-slip stress regime, contains 3

#### **Figure 11.**

*Fracture propagation in the presence of a natural fracture flaw (in red) close to the vertical path from the injection point in specimen-2. Stiff layer (dark blue) bounded by soft layers (light blue) in the presence of a (top) perfect interfaces (case1, top), and weak interfaces (case 2, bottom). Notice the step-over occurred in the weak interface but not in the perfect interface. The animated figure can be seen in Video 1 available from (can be viewed at) https://youtu.be/oqDx96YXSvQ* 

#### **Figure 12.**

*(a) Experimental setup and laser scan showing final 3D fracture geometry (green) arrest against 2 natural fracture planes in [27] laboratory test (b) 3D ADaM MPM result showing the fracture geometry and its arrest against the dipping fracture planes represented as weakly bonded interface. Animated version of the 3D MPM result can be seen in Video 2 available from (can be viewed at) https://youtu.be/jdfDAM2qi-8.* 

*Surface Drilling Data for Constrained Hydraulic Fracturing and Fast Reservoir Simulation… DOI: http://dx.doi.org/10.5772/intechopen.84759* 

 natural fractures with a 40° strike relative to SHmax and 75° dip. In the experiment, the pressurization was accomplished by eight perforations, four in each side, parallel to the strike of the fractures. The natural fractures are not mineralized, so they were modeled with the Coulomb frictional contact law with μ = 0.85 according to the laboratory experiment. **Figure 12** shows the experimental results vs. the 3D MPM numerical model. The geomechanical modeling tool was able to reproduce the main features, especially the turning of the hydraulic fracture and arresting of the fracture by the nearest natural fractures. No crossing of natural fracture was observed neither in the numerical results nor in the experiment. This result highlights the ability of the combined use of the fully 3D ADaM model with the interface modeling tools in capturing the interaction between hydraulic and natural fractures in field conditions requiring full 3D geomechanical modeling tools.

The results from these decoupled processes, that attempt to quantify the geomechanical impact of geologic characteristics, can be used to constrain a 3D grid based planar hydraulic fracturing simulator that will include proppant transport.
