Effects of fracturing parameters on fracture network evolution during multicluster fracturing in a heterogeneous reservoir

There are some major problems in the study of fracture network evolution during multicluster fracturing, such as the simple assumption that the reservoir is homogeneous, without considering the heterogeneity of rock mineral particles. In addition, different viscosities used in fracturing fluid and fluid injection methods can affect fracture propagation. As reservoir stimulation enters the 2.0 era, it is important to study the evolution of multicluster fracture networks based on reservoir heterogeneity to improve the stimulated reservoir volume. For this purpose, a 2D, coupled stress-seepage-damage field multicluster fracturing numerical model was developed in this study, and a globally embedded 0-thickness cohesive element was used to simulate the random propagation of hydraulic fractures. Moreover, the elastic modulus, tensile strength and mixed fracture energy of the cohesive element mesh were randomly assigned using the Weibull distribution probability density function to simulate the reservoir heterogeneity. In addition, the dynamic distribution of the injected fluid rate during multicluster fracturing was implemented based on the Bernoulli equation. The capability of the model was validated with analytical solutions of the Khristianovich-Geertsma-de Klerk (KGD) problem and similar numerical modeling criteria of multicluster fracturing and was used to investigate the geological and engineering factors that influence the evolution of the fracture network. The results show that the fracture network formed by hydraulic fracturing is more complex when the reservoir has significant heterogeneity and the injection pressure has a larger extreme value. As the reservoir becomes increasingly homogeneous, using low-viscosity fracturing fluid does not significantly contribute to fracture network complexity or effectively reduce the extreme value of the injection pressure. Shortening the pad fluid injection time can significantly increase the fracture network complexity, but it increases the extreme value of the injection pressure, which requires a higher pressure tolerance for fracturing equipment. The results of this study have important significance for the parameter optimization of multicluster fracturing.

Automated summary

There are several problems in studying the evolution of fracture network during multicluster fracturing, such as assuming the reservoir is homogeneous and neglecting the heterogeneity of rock particles. The study developed a 2D numerical model using a coupled stress-seepage-damage field approach to simulate hydraulic fracture propagation. The model considered reservoir heterogeneity by assigning randomly distributed properties to cohesive elements. The results showed that fracture network complexity increases with reservoir heterogeneity and injection pressure. Low-viscosity fracturing fluid does not significantly impact fracture network complexity. Shortening pad fluid injection time increases fracture network complexity but also raises the extreme value of injection pressure.