The main goal of this project is to make efficient and accurate predictions on production patterns using innovative flow simulators.
In carbonate reservoirs, modelling of fractures constitute a vital ingredient for useful predictions on well productivity and field-scale production. In the case of complicated fracture orientation and connectivity, dual-porosity models might not actually reflect reality, and the use of a Discrete Fracture Matrix (DFM) model, which explicitly represents the fracture geometry, is preferred. However, the main problem with DFM models, especially in the case of complex realistic fracture networks, is the gridding step. One challenge, therefore, has been to improve the fracture discretization approach (realism of our flow models) while keeping computational complexity at a similar level.
Naturally fractured reservoirs (NFR) are highly heterogeneous and contain large uncertainty in its input parameters, which in turn results in huge uncertainty in the flow response. This requires the need for uncertainty quantification. Performing uncertainty quantification on a large (high-fidelity) ensemble (i.e., a set of reservoir models) can become computationally unfeasible. This has led to the development of several techniques that try to speed up the uncertainty quantification process and hence decision-making process while keeping a certain level of accuracy.