High Resolution Discrete Fracture Network Application for Polymer Flooding Design in a Naturally Fractured Carbonate

Despite their higher complexity (Juri et al., 2015) and usually more challenging commercial development, naturally fractured reservoirs account for a significant portion of oil and gas reserves worldwide (Sun et al., 2021). Typically, natural fractures tend to enhance the productivity of the wells, yet they also tend to accelerate reservoir depletion, often leading to sub-optimal field production and leaving significant volumes of hydrocarbons behind (Aguilera, 1995). In this work, we propose a specific polymer injection design that can provide the conditions for fracture-matrix counter-current flow to develop in a naturally fractured carbonate reservoir. In turn, this flow could trigger a virtuous cycle where the displacement front is progressively slowed down, increasing the efficiency of the displacement process and the oil recovery. This study focused on the integration of multiple sets of data to characterize karstic and tectonic fractures in a discrete fracture network (DFN) model and its posterior use in a dual medium simulation model to determine polymer flooding optimal spacing and injection strategy in a complex, naturally fractured carbonate system. An innovative and integrated approach combining 3D seismic data, bore-hole imagery (BHI), cores, and production data was applied to characterize and represent karstic features. The applied workflow consisted of (1) identification and manual picking of karstic features on BHI, (2) deterministic picking of karstic features as geobodies on the 3D seismic (enhanced similarity volume), (3) integrated implementation of the karstic features into the geological model using advanced geostatistical methods (Multi-Points Simulation, or MPS), and (4) implementation of resulting enhanced reservoir properties on a fit for purpose high-resolution dynamic model (dual porosity/dual permeability). Multiple simulations were run to evaluate different sensitivities including injection rates, injection strategy, completion approach, and producer-injector pattern spacing. Particularly for the latter, a robust karst/fracture system characterization was critical to propose optimal pattern sizes which aim to simultaneously avoid early polymer breakthrough -in shorter than optimal designs and minimize potential shear thickening degradation effects tied to higher polymer throughput required by excessive producer-injector distancing. In terms of the completion interval, the DFN-derived properties were also strongly conditioning the selection of the injection interval with noticeable effects and contrasting results. Because of the superposed features constituting the total fracture system and their different origins, a field-level comprehension of anisotropy and local intensity of the fractures is critical for selecting both the wells for the injectivity test and the potential area for the pilot in the next stage of the project.