Numerical Investigation and Optimization of Multiple Fractures in Tight Gas Reservoirs

M. Z. Hou,L. Zhou
2013-01-01
Abstract:The main objective of this article is to investigate the influence of individual fractures on each other in a multi-fracture system, including space optimization, by using the FDM program FLAC3D(plus) featuring own developments, which treats all fractures in a single 3D geometric model, under a 3D stress state taking hydro-mechanical coupling effects into consideration. A case study was conducted on a horizontal wellbore at location A, which was simulated hydraulically with a total of eight transverse fractures in summer 2009. Transverse multiple fractures were simulated using the modified continuum method. In the simulation all fractures were generated in one single model, comprising 22 different rock layers. Each layer was assumed to be homogeneous with regard to its rock and hydro-mechanical parameters. Thus the influence of the individual fractures on each other can be investigated. The simulation procedure applied, which is a consecutive execution of a hydraulic and a mechanical computation, is the same for all fractures. The only differences are the primary in-situ stresses, the initial pore pressure, the injection parameters (location, rate, volume, duration), which lead to different patterns of fracture propagations. But there are still some common points, such as irregular patterns of the fracture front, which represents the heterogeneity of the model. All fractures (I to 8) have their fracture average half-length between 70 in and 115 in, height 93 in to 114 m and average width 18 mm to 31 mm. The percentage difference of fracture height for individual fractures is obviously smaller than that of the fracture half: lengths, because the fracture barriers at the bottom and especially at the top of a reservoir limit the fracture propagation in z-direction. In comparison with the semi-analytical simulator (FracPro) most results match well. Simulation of multiple fractures at location A, with the newly developed algorithms, shows that individual transverse multiple fractures at distances between 100-150 m will certainly have an influence on each other in terms of the stress change, but not pore pressure. Generally this influence as simulated at this location has a positive effect on the fracture geometry as well, and may lead to a bigger half-length and a smaller width of fracture. In order to maximize the production rate at this specific site, transverse multiple fractures with an optima fracture distance of ca. 57 m should be implemented. This optimal fracture spacing is derived from the contour of 5 degrees direction change of the primary maximum horizontal stress, to ensure that only transverse fractures are generated at this location. However, the influence on longitudinal multiple fractures is very obvious, as observed from the numerical simulations of longitudinal multiple fractures at location B. Under the circumstances, it is recommended that the boreholes should not be drilled exactly in the direction of maximal horizontal stress (+/- 15 degrees), in order to avoid the overlapping of fractures.
What problem does this paper attempt to address?