Open Access Unrestricted online access star

This synthesis highlights innovations addressing reservoir heterogeneity and fracture dynamics through integrated numerical modeling, data assimilation, and multi-physics coupling. Ensemble-based algorithms (e.g., ES-MDA) enhance history matching by assimilating 4D seismic and production data, reducing uncertainties by 15–20%. Hydro-mechanical models optimized with true triaxial experiments guide Discrete Fracture Network (DFN)-driven hydraulic fracturing, boosting shale gas productivity by 40%. Proxy models like INSIM-FT and Physics-Informed Neural Networks (PINNs) enable rapid simulation, cutting computational time from weeks to hours while maintaining >85% accuracy. Machine learning (XGBoost) achieves 92% permeability prediction in carbonates, while dynamic heterogeneity analysis reveals fracture-induced permeability contrasts exceeding 103. Geomechanical frameworks quantify risks in salt cavern storage (0.12% annual creep strain) and fractured reservoirs, extending operational lifespans by 20%. Field applications demonstrate 8% recovery gains in carbonate fields via 4D seismic integration and 60% leakage risk reduction through multi-physics cement design. Emerging trends fuse data-physics models (30–50% efficiency gains) and cross-scale simulations, while challenges persist in proppant transport modeling and sparse 4D data. Future directions prioritize quantum computing for fracture networks, IoT-enabled digital twins, and adapting reservoir engineering to carbon sequestration, positioning the field as pivotal for sustainable energy transition.

Creative Commons Licence cc by cc https://creativecommons.org/licenses/by/4.0/

eng

Freely available e-book