Enhancing Understanding of Hydraulic Fracture Tip Advancement through Inversion of Low-Frequency Distributed Acoustic Sensing Data

Characterizing the fluid-driven fracture tip advancing process presents a significant challenge due to the difficulty of replicating real-world conditions in laboratory experiments and the lack of precise field measurements. However, recent advances in low-frequency distributed acoustic sensing (LF-DAS) technology offer new opportunities to investigate the dynamics of propagating hydraulic fractures. In this study, we propose an iterative inversion method to characterize fracture-tip advancing behaviors using LF-DAS data. A forward geomechanical model is developed using the three-dimensional displacement discontinuity method, and the optimization is realized by a conjugate gradient method. The performance of the inversion algorithm is demonstrated using a synthetic case, in which the fracture half-length evolution and propagation velocity match well with the reference solutions. Additionally, the averaged fracture cross-section area, fracture volume, and fracturing fluid efficiency can also be estimated, showing good agreements with true values of the synthetic case under reasonable assumptions. Then a field case with a single-cluster hydraulic fracturing treatment from the Hydraulic Fracturing Test Site 2 project (HFTS2) is studied. Our analysis of the inversion results reveal that the fracture propagates intermittently, as evidenced by the fracture half-length evolution. This unique field evidence can guide modeling efforts to incorporate this important physical behavior into fracture models, and the secondary information gathered from the study, including fracture cross-section area and volume, can help evaluate and optimize fracturing efficiency.