| As part of a larger effort in developing predictive models for the Elm Coulee Bakken Formation, this thesis demonstrates the use of discrete fracture networks (DFNs) to characterize natural fractures and understand production variability in the Bakken Formation. Historically, reservoir simulation has used transmissibility multipliers to simulate enhanced fracture permeability. Not only does this strategy struggle to match the high early production followed by the long, flat decline production life typifying unconventional reservoirs, but traditional simulation techniques do not account for fracture volumetric changes intrinsic to pressure dependent permeability.;However, with the advent of DFN modeling tools and increased understanding of unconventional reservoirs, a more refined approach may be viable. To compare the effectiveness of DFN models with traditional fracture modeling of natural fractures using permeability multipliers, a study area comprised of nine wells producing from the Elm Coulee Bakken Formation was selected to evaluate the two techniques. DFN parameters were based on local geologic history and observed fractures in cores. Study area production performance was history matched as a heuristic of simulation success using both the traditional simulation strategy and the DFN approach. Comparing outcomes of both techniques will determine whether DFNs are a viable method for describing natural fractures in the Elm Coulee Bakken. If so, using DFNs will lead to improved predictive reservoir modeling techniques.;Keywords: Bakken, Natural Fractures, Elm Coulee, Reservoir Simulation, Unconventional, Discrete Fracture Network. |
