An Improved Technique for Predicting Vibration Levels from Tunnel Blasting

Despite increasing competition from mechanical methods of tunnelling, the drill and blast method is often still the most viable method of excavating tunnels in strong and abrasive rock. To advance a tunnel using drill and blast, explosives are loaded into boreholes in the rock and detonated according to a prearranged sequence. On detonation of the explosive, the rock is fractured and displaced from its original position, leaving behind the desired void. Among the secondary effects of a tunnel blast are ground vibrations caused by elastic disturbances that propagate away from the blasted tunnel face. The possibility that such ground vibrations may cause permanent damage to property and substantial nuisance has generated significant opposition to the use of drill and blast, particularly in urban areas. It is therefore essential that the blasting engineer is able to predict Peak Particle Velocity (PPV) vibration levels at locations in the vicinity of the blast site. Such predictive capability enables the engineer to design blasts so as to ensure that ground vibrations can be kept within acceptable levels. The objective of this project was to analyse data obtained from the ground vibration monitoring program undertaken during the drill and blast excavation of a hard rock tunnel in Central England. Seismograph recordings of tunnel blasts contain a vast amount of information that cannot possibly be expressed using a single PPV value. The main focus of the study was to investigate whether predictive capability could be improved by using scaled distance models derived for selected PPV components relating to specific detonator timings in a given tunnel round. This study has established a method of analysis that can unlock this information for predictive purposes. By permitting the extraction of sub-event Peak Particle Velocity (PPV) values and allowing Scaled Distance (SD) models to be derived for individual period numbers or groups of period numbers, the method has the potential to vastly improve PPV prediction from tunnel blasting. The method established could also help diagnose blasting problems and help target blast design modifications, resulting in improved efficiency and excavation rates.