Mobility control of gas injection in highly heterogeneous and naturally fractured reservoirs

Since a significant portion of the world's oil reserves resides in naturally fractured reservoirs (NFR), it is important to maximize oil production from these reservoirs. Mobility control EOR techniques, such as water alternating gas (WAG) and foam injection, may be used in NFRs to improve oil recovery. Foam injection may be modeled by empirical or mechanistic models, the latter being capable of representing foam generation and coalescence effects. Numerical models are needed to evaluate EOR techniques in NFR. The Embedded Discrete Fracture Model (EDFM) is capable of representing conductive faults or fractures and describing NFR and unconventional reservoirs as a triple porosity medium (hydraulic fractures, natural fractures, and matrix). This work aims at developing a general EDFM framework to allow the evaluation of different mobility control EOR methods in NFR. The mobility control EOR methods evaluated were the WAG and continuous foam injection. The formulation used to evaluate mobility control by foam injection in NFR was the population balance assuming local equilibrium and the Pc* models. Nanoparticle transport models (Two Site and Two Rate models) were implemented and validated to allow simulation of nanoparticle stabilized foam injection. An EDFM preprocessor was further developed and validated against the in-house fully implicit simulator, unstructured grid models from the literature and fine-grid models using a commercial simulator. Simulation run time was reduced by applying a porosity cut-off in the fracture cells assuming constant fracture conductivity. Validation case studies included multi-fractured wells producing through depletion and a 2D quarter five-spot production scheme (water and miscible gas injection) in NFR. We obtained a good agreement between EDFM, unstructured grid, and fine-grid models. Application case studies included 3D models under water, miscible gas and WAG injection, which confirmed the efficiency of the EDFM in modeling complex fracture networks. We used the EDFM to simulate multilateral well stimulation and we performed an automated history matching of the production data of a field test. The foam model and the nanoparticle transport models were validated against experimental data from the literature. It is concluded that the effect of fractures on hydrocarbon production depends on fracture network connectivity, which may be modeled using the EDFM preprocessor. Simulation results using mobility control EOR methods show considerable improvements in oil recovery due to a postponement in gas breakthrough.