DEM‑Based Hydraulic Fracturing Simulation in Heterogeneous Formations: Joint Validation with Acoustic Emission and Active Wave Propagation
Yao Zhu1, Chun Liu1, Jianbo Wang2, Yuhang Lin1, Wenqiang Xia1, Baojun Wang1
Abstract: The mechanis m of hydraulic fractures propagation and interaction is not well understood. Based on on-site experimental data, this paper uses the Discrete Element Method to simulate the hydraulic fracturing process in heterogeneous rock formations, to correlate macroscopic responses (mechanic and acoustic responses) to the contact breakages. The rock model exhibits a brittle failure characterized by injection pressure drops due to cracks. The fracture development is monitored acoustically by measuring both the passive acoustic emission (AE) and the velocity change in the active wave propagation. We have identified cracks as AE sources, and the differences in crack formation result in a dual-frequency characteristic in the AE signals: the low-frequency component is associated with tensile cracks with larger opening, while the high-frequency component corre sponds to tension-shear micro-cracks with s maller opening. The location of cracks can be determined by the time-differences an alysis. The b-value of AE signals in the Gutenberg-Richter law significantly decreases before the peak injection pressure. Studies of active wave propagation reveal the reduction in P-wave velocity and enhanced coda wave scattering caused by the anisotropic hydraulic fractures. The reduction in P-wave velocity is closely related to the macroscopic stiffness degradation caused by cracks, while correlation a nalysis effectively distinguishes local fracture intrusion from remote stress disturbances. By combining DEM with acoustic monitoring, this paper intends to provide a comprehensive fra mework for understanding hydraulic fracture evolution and enhancing fracture detection and characterization.
Keywords: Discrete element method; MatDEM; Hydraulic fracturing; Acoustic emission; Wave propagation

Fig.2 (a) The basic geotechnical solid consists of discrete elements (The color represents the element size); (b) the pore network based on Discrete Element Model; (c) the model including elements and pores, where the fluid within a single pore is uniformly distributed; (d) Seep age in the pore networks; (e) schematic diagra m of the hydraulic radius (the orange circles represent real discrete element elements, and the blue dashed circles represent elements with hydraulic radii. The dashed rectangle between the two elements is the pore throat with length l and width dw); (f) Elements undergo displacement and pore structure changes under fluid–solid interaction (color figure online)

Fig.3 (a) Schematic diagra m of fracture fluid invasion; (b) the connection is broken under the action of high-pressure fluid and generates acoustic emission signals

Fig.5(a) Schematic diagra m of the site model; (b) application of lateral in-situ stress and the micro-cracks generated after cutting the prefab ricated fracture (color represents particle size); (c) colorful elements around the model represent artificially added serrated high-da mping layers to simulate a semi-infinite space; (d) initial pore pressure distri bution (color figure online)

Fig.6 (a) On-site observation of hydraulic fractures developed along bedding planes; (b) on-site observation of main fracture intersecting with bedding planes; (c) another on-site observation of the main frac ture intersecting with bedding planes; (d) the final hydraulic fracture development in the simulation results, where the black short lines represent the crack generated by the fracturing; (e) variation of on-site injection pressure; (f) variation of simulated injection pressure (color f igure online)




Zhu Y, Liu C, Wang J, et al. DEM-Based Hydraulic Fracturing Simulation in Heterogeneous Formations: Joint Validation with Acoustic Emission and Active Wave Propagation[J]. Rock Mechanics and Rock Engineering, 2025: 1-21.