Evaluation of proppant injection based on a data-driven approach integrating numerical and ensemble learning models

Injecting proppant to prop open fluid-driven fractures in subsurface reservoirs is one of the key missions hydraulic fracturing. However, quantitative evaluation distribution successfully propped limited due infeasibility direct measurement. This work defines an indexing parameter for field practice estimate proportion proppant-filled reservoir – filling index (PFI). A new data-driven workflow, combining numerical models and ensemble learning algorithm, proposed trained on records both screen-out near cases then applied predict PFIs regular cases. The algorithm performance promoted via variable importance measure (VIM) analyses a backward elimination strategy. Four twelve are presented demonstrate predicted PFI highlight its potential utilizations. workflow evaluate injection quantitatively reveal any mismatch between underground fractures, which may be essential post-fracturing characterization improve oil & gas recovery, sequestration CO 2 , storage recovery H deep geothermal fluids as important components enabling energy transition. • An (PFI, fractures) defined. reveals reservoirs. built practice. Proppant evaluated analysis. Fracture evolution qualitatively described characterizing fractured