Realistic representation of Discrete Fracture Networks (DFN) on field scale simulation models is computationally infeasible due to extreme aspect ratio of fracture aperture to length. Our modeling approach entails developing effective flow characteristics of discrete fractures at micro and macrofracture scales without explicitly representing the fractures on a grid. A random walker simulation is used that moves walkers along implicit fractures honouring the intersection characteristics of the fracture network. Several superimposed realizations of micro and macrofracture networks enable us to capture the uncertainty in the network. For different DFN models, variation in permeability is indicated by variance of travel time and percolation statistics indicative of network connectivity. The effective random walker statistics corresponding to microfractures are calculated, and subsequently specified as equivalent matrix properties for the macrofracture simulation. The random walker procedure is applied on the macrofracture network with the effective permeability from the microfracture network specified as the matrix permeability. Comparison of random walker results to high-resolution flow simulation for a few simple networks is shown.
The method was applied for a producing naturally fractured carbonate reservoir. Field survey identified six different fracture sets. Seismic data was processed to obtain fracture intensity maps. Taking these as input, DFN models were created and permeability upscaling was performed to effectively characterize the fracture distribution and study its impact on flow. Maps of effective fracture conductivity and matrix to fracture transfer coefficients were generated. Permeability anisotropy due to orientation of fracture sets was also quantified. An important observation was the definition of percolation threshold for each direction. It defines the minimum fracture intensity required for the network to percolate and hence affects the effective grid-block permeability. |
