The main challenges in modeling fluid flow through naturally-fractured carbonate karst reservoirs are how to address various flow physics in complex geological architectures due to the presence of vugs and caves which are connected via fracture networks at multiple scales. In this paper, we present a unified multi-physics model that adapts to the complex flow regime through naturally-fractured carbonate karst reservoirs. This approach generalizes Stokes-Brinkman model (Popov et al. 2007). The fracture networks provide the essential connection between the caves in carbonate karst reservoirs. It is thus very important to resolve the flow in fracture network and the interaction between fractures and caves to better understand the complex flow behavior. The idea is to use Stokes-Brinkman model to represent flow through rock matrix, void caves as well as intermediate flows in very high permeability regions and to use an idea similar to discrete fracture network model to represent flow in fracture network. Consequently, various numerical solution strategies can be efficiently applied to greatly improve the computational efficiency in flow simulations.
We have applied this unified multi-physics model as a fine-scale flow solver in scale-up computations. Both local and global scale-up are considered. It is found that global scale-up has much more accurate than local scale-up. Global scale-up requires the solution of global flow problems on fine grid, which generally is computationally expensive. The proposed model has the ability to deal with large number of fractures and caves, which facilitate the application of Stokes-Brinkman model in global scale-up computation.
The proposed model flexibly adapts to the different flow physics in naturally-fractured carbonate karst reservoirs in a simple and effective way. It certainly extends modeling and predicting capability in efficient development of this important type of reservoir.
Naturally fractured carbonate karst reservoirs are composed of porous materials; while at the same time will contain relatively large void spaces in the form of fractures, small cavities, and caves, which are called vugs in geological literature. These vugs and caves are typically interconnected via fractures at multiple scales. This presents a major challenge in modeling the fluid flow through such formations because of the co-existence of porous and free-flow regions on multiple scales. The presence of individual void spaces such as vugs and caves in a surrounding porous media can significantly alter the effective permeabilities of the media. Furthermore, interconnecting fractures can form various types of connected networks which change the effective permeability of the media by orders of magnitude. An additional factor which complicates the numerical modeling of such systems is the lack of precise knowledge on the exact position of the interface between the porous media and void spaces. Finally, the effects of cave/fractures that are filled in by loose material (sand, mud, gravel, etc), the presence of the damage at the interface between porous media and vugs/caves and the roughness of fractures can play very important role in the overall response of such a reservoir.