Blast Vibration Prediction In Hard Rock Underground Mines

Abstract

Blast-induced ground vibration is one of the most significant environmental and safety hazards associated with 
drilling and blasting operations in hard rock underground mines, with the potential to damage adjacent 
excavations, support systems, and surface structures, while also endangering personnel and equipment. Over the 
past five decades, researchers have proposed a wide spectrum of predictive approaches ranging from classical 
empirical scaled-distance equations to sophisticated statistical, soft-computing, and hybrid artificial intelligence 
models aimed at forecasting peak particle velocity (PPV) with greater accuracy. This review paper undertakes a 
systematic meta-analysis of past research on blast vibration prediction, tracing the evolution of predictor 
equations from the United States Bureau of Mines formula through Langefors-Kihlstrom, Ambraseys-Hendron, 
and Indian Standard models, to contemporary machine learning structures including artificial neural networks, 
support vector machines, adaptive neuro-fuzzy inference systems, and hybrid metaheuristic-optimized models. 
The paper synthesizes findings from over one hundred studies, critically evaluates the strengths, limitations, and 
predictive accuracy of each modelling category, and identifies the dominant controllable and uncontrollable blast 
design parameters influencing vibration intensity. Gaps in existing literature, including limited generalizability 
across geological settings, insufficient real-time monitoring integration, and inadequate consideration of 
underground confinement effects, are highlighted. The review concludes that while data-driven and hybrid models 
consistently outperform traditional empirical equations in predictive accuracy, site-specific calibration, larger 
standardized datasets, and integration of geotechnical uncertainty remain essential future research directions for 
reliable blast vibration management in underground hard rock mining environments.