Workflow for well index calibration using DFN modeling in the near-well region of fractured reservoirs

Conventional upscaling techniques, such as the Oda method, combined with dual-porosity models, often overestimate near-well permeability and fail to reproduce discrete fracture network (DFN) connectivity. This work proposes a cost-efficient methodology to calibrate the well index (WI) and mitigate this bias using explicit DFN modeling in the near-wellbore region. The WI is obtained from a DFN-based simulation (DKDFN) executed only for the first time step and subsequently replaces the conventional WI in the dual-permeability (DK) model, yielding a hybrid formulation termed DKH. Automated scripts allow the workflow to be applied to ensemble-based studies under uncertainty. The methodology was evaluated in two cases: (i) a validation model with an idealized DFN and (ii) a real Brazilian pre-salt field. An adapted Normalized Quadratic Deviation with Sign (NQDS) metric quantified the bias between DK and DKH. In the validation case, DKH reproduced the reference behavior for highly connected fractures and avoided overestimated permeability in poorly connected DFN. Field-scale results revealed notable differences between DK and DKH production responses. A correlation-matrix heatmap indicates that interference from fractured wells can increase NQDS values in non-fractured wells. The proposed workflow improves WI calibration by incorporating near-well DFN geometry without increasing simulation time.