School of Civil and Environmental Engineering

Ph.D. Thesis Defense Announcement

Experimental and Numerical Modeling of Fluid Injection into Unconsolidated Formations

By

Longde Jin

Advisor:

Dr. Haiying Huang (CEE)

Committee Members: Dr. Susan Burns (CEE), Dr. David Frost (CEE),

                                                      Dr. Leonid Germanovich (CEE), Dr. Ying Zhang (ECE)

Date and Time: Friday, August 18th, 2017 at 10:00 am

Location: Sustainable Education Building (SEB) 122

Understanding the failure mechanisms and fluid flow patterns in response to fluid injection in unconsolidated formations is critical to reservoir stimulation and production. While hydraulic fracturing is commonly employed to create pathways of increased permeability to connect a wellbore with the reservoir matrix, the fracturing process in such weakly cemented and highly permeable formations remains poorly understood. Meanwhile, production in these reservoirs is often accompanied by fine migration, which could substantially reduce the fracture permeability and therefore hamper the productivity in the long term. In practice, acid could be injected into the production zone in order to recover the fracture conductivity.

Both the fracturing and acid cleaning processes in the unconsolidated reservoirs are examined in this work. Specifically, we focus on characterizing how the fracture morphology and fluid leakoff are affected by the injection rate, fluid rheology and formation permeability. A series of injection experiments is performed with mixtures of sand and silica flour as analog materials. We show that as the weight percentage of the silica flour increases, the matrix permeability decreases significantly and the capillary effect becomes non-negligible. Compared with the injection experiments with pure sand, an additional dimensionless number incorporating surface tension needs to be introduced in order to characterize the fluid-grain displacement process. Effect of the non-Newtonian fluid rheology is subsequently analyzed using the discrete element method coupled with a pore-network model. A quasi-steady-state fluid flow algorithm is developed to enhance numerical stability and computational efficiency in modeling fluid injection in a wellbore. It is shown that the high shear rate rheology is critical to the near-wellbore failure and fluid flow. Furthermore, a hybrid phase field method is constructed to model the fracturing process. The benefit of the phase field method is that creation of a fracture could be modeled through explicit consideration of porosity change. Finally, the acid cleaning process is simulated using a hydro-chemically coupled scheme implemented in an equation based solver. Effects of the injection rate, the acid reaction rate and the fracture conductivity are examined. Results from this numerical analysis could serve as guidelines to optimize the field practice.