Numerical study of the stimulation related thermo-hydro-mechanical processes in tight gas and deep geothermal reservoirs

In recent times, the efficient development of unconventional gas/oil reservoirs and deep geothermal resources has become a focal point of research in the field of energy. Since the inception of stimulation in the 1940s, the technique has been revolutionized especially by combining hydraulic fracturing with horizontal well technology. Hydraulic fracturing is a complicated operation owing to several important issues, such as fluid viscosity’s influence on the stimulation results and the induced seismicity during the fracturing operation or even reactivation of the natural faults etc. In this thesis, targeted improvements have been achieved through the development of a series of mathematical/physical models, and their implementation into the existing numerical tools (FLAC3Dplus and TOUGH2MP-FLAC3D), including: (a) a new thermal module for FLAC3Dplus based entirely on the finite volume method (FVM), which is especially developed for the fracturing process and can also achieve the modeling of gel breaking; (b) a rock damage module of TOUGH2MP-FLAC3D, which also considers the impacts of rock damaging process on evolution of permeability; (c) an in-depth improved FLAC3Dplus simulator that obtains the ability to simulate a 3D fracture propagation with arbitrary orientation. After the corresponding verifications of the improved tools, different case studies are conducted and analyzed. The following conclusions can be drawn from these case studies: 1) Systematic study of the fluid viscosity’s impacts on shaping of a fracture in tight sandstone: a) the fracture’s growth during stimulation is governed by two competing energy dissipation mechanisms (viscous flow and fracturation) and two competing storage mechanisms (in the fracture or in the porous matrix). The system tends to be storage (in fracture)-dominated, when fracture’s leak off ability is sufficiently low (depending on the combined effects of formation permeability and fluid viscosity); b) Change in horizontal stress (σh) and pore pressure (Pp) are actually two competing mechanisms. Change of the pore pressure is mainly driven by the leak off process but the minimum horizontal stress varies mainly due to the fracture pressure, i.e. σh alters more intensively when a system is hard to leak off. Furthermore, the fluid viscosity also affects the final shape of fracture through its proppant-carrying ability. To make the simulation results closer to the practical situation, the function of simulating gel breaking was developed for the thermal module. Through conducting a real case study in tight gas reservoir Leer (considering the THM coupling effects and the gel breaking process), it was found that not only the leak off but also the proppant’s settling would be accelerated by a smoother fluid. A possible way to solve this problem is introducing gas-based fracturing such as supercritical CO2, as in this method the fracture’s close rate can be much faster than the settling. 2) Geothermal utilization induced microseismic in stimulation and production phase of Landau project: a) Core area within and around the natural faults is more susceptible to the stimulation work. Its plastic strain, therefore, increases further even after the propagation has stopped; b) Mechanic and hydraulic equilibrium cannot be achieved immediately. Thus in the post failure process, fracturing and seismicity occurred even after the injection is stopped; c) intense rise or fall in injection/production rates induces disturbance in the system. A huge difference between them possesses the same effect. The stronger the disturbance, the more intensive the fracturing and seismic would be. It can be concluded that, immoderate changes of injection/production rate or a huge difference between them lead to induced seismicity, while the reactivation trend of natural faults is impacted by the absolute value of injection/production rate, i.e. intensity of the operation. A reasonable response to reduce the risks is: reducing the injection and production rates immediately with a moderate and equal rate (injection/production) when critical seismicity magnitude (e.g. ML≥2.0) occurs. 3) Advanced FLAC3Dplus: this powerful simulator has gained new features through several in-depth improvements. After implementing the triangle prism element and reprograming the mechanic and hydraulic codes, this simulator is verified to be able to model the fracture propagation with arbitrary orientation. More importantly, it overcomes the shortcoming of XFEM (extended finite element method) and can be applied in 3D situation.