Real-time geological hazard detection ahead of tunnel face using blast-induced seismic signals: A case study in Tashan Tunnel, China

Surrounding rock in near-surface tunnels of Yantai city, China, characterized by high weathering, densely developed fractures, and locally broken rock masses, resulting in challenges such as poor stability and low excavation efficiency. However, existing geophysical detection methods generally require suspension of construction activities, affecting tunneling progress. This study develops a novel method utilizing blast-induced seismic signals during tunneling for advance prediction, achieving high-resolution real-time detection of fractured zones without disrupting ongoing construction activities. Based on field data from the Tashan Tunnel in Yantai City, complete ensemble empirical mode decomposition (CEEMD) was applied to analyze blast-induced signals, thereby guiding parameter selection for three-dimensional wavefield simulation. The proposed method employs seismic interferometry to reconstruct simulated blast signals into virtual shot gathers equivalent to active source excitation, accurately predicting shear wave velocity variation interfaces (lithological boundaries) ahead of the tunnel face with minimal error (approximately 0.3 m). Field implementation in the Tashan Tunnel Project demonstrated the method's effectiveness, successfully predicting densely fractured zones more than 30 m ahead of the excavation face with high accuracy (error of about 1 m). This prediction guided timely implementation of preventive measures, including face hazard removal, prompt closure of primary support, and short-distance advance drilling to assess whether modifications to excavation methods were necessary. This innovative approach employs random blast signals for risk detection, providing valuable reference for optimizing advanced support systems and excavation methods in tunnel projects through complex geological environments.