Feature identification in multi-cluster true triaxial hydraulic fracturing using integrated fiber optic and acoustic emission monitoring

Abstract

Hydraulic fracturing test sites have been widely established to investigate fracture propagation, often integrating techniques such as distributed fiber optic sensing, microseismic monitoring, and core sampling. However, these multi-source monitoring approaches are costly and time-consuming to implement. Considering the similarity of monitoring principles, single/multi-cluster true triaxial hydraulic fracturing simulations were conducted on continental shale outcrops using a combined system of 16-channel acoustic emission and distributed strain fiber optics. Fracture propagation was controlled by both tensile and shear stresses, with shear contributions typically below 15 %, and was predominantly observed along lamination fractures activated at later stages. Competitive interactions among multi-cluster fractures were observed, reflected in variations in initiation time, sequence, and propagation paths. The stress shadows between two adjacent fractures can overlap, impeding inward propagation and resulting in a characteristic “back-to-back” fracture growth pattern. Simultaneously, stress shadow effects may lead to inhibited initiation in adjacent clusters, enhance fracture linkage along bedding planes, and greater deviation from the S_Hmax direction. Strain rate responses recorded by adjacent well fiber optics exhibited characteristic geometries: heart-shaped patterns for single fractures and V-shaped patterns for two adjacent fractures, with strain rates ranging from 0 to 300 με/s and durations of 2–4 s. As the fracture approached the monitoring fiber, the strain signature evolved from arcuate (approaching), to heart-shaped (contact), and spindle-shaped (passing through), corresponding to increasing fracture proximity. The findings provide practical guidance for interpreting field monitoring data and optimizing stimulation designs in hydraulic fracturing operations.