Intelligent Approach in Smart Microchip Proppants Data Processing for Complex Hydraulic Fractures Diagnostic

Unconventional oil and gas reservoir development requires an understanding of the geometry and complexity of hydraulic fractures. The current categories of fracture diagnostic approaches include methods for near-wellbore (production and temperature logs, tracers, borehole imaging) and far-field techniques (micro-seismic fracture mapping). These techniques provide an indirect and/or interpreted fracture geometry. Therefore, none of these methods consistently provides a fully detailed and accurate description of the character of created hydraulic fractures. This study proposes a novel approach that uses direct data from the injected fine size and battery-less Smart MicroChip Proppants (SMPs) to map the fracture geometry. This novel approach enables direct, fast, and smart of the received high-resolution geo-sensor data from the SMPs collected in high pressure and high-temperature environment and maps the fracture network using the proposed Intelligent and Integrated Fracture Diagnostic Platform (IFDP), which is a closed-loop architecture and is based on multi-dimensional projection, unsupervised clustering, and surface reconstruction. Affine transformation and a deep ANN are integrated to control the stochasticity of clustering. IFDP proves its efficacy in fracture diagnostics for 3 in-house design synthetic fracture networks, with 100% consistency, between 75-100% in prediction capability, and between 80-100% in execution robustness. The integration of affine transformation firmly increases the performance of unsupervised clustering in IFDP. The low bounds of prediction capability (75%) and execution robustness (80%) scores are caused by the "joint" locations in the synthetic fracture network with the highest complexity.