Comparison of Excavation Disturbance Around Deep Tunnels in Hard Rock Using Acoustic Emission and Ultrasonic Velocity Methods Acoustic Emission and Ultrasonic Velocity Methods

Abstract

Acoustic emission (AE) and ultrasonic velocity monitoring studies have been undertaken at both the Atomic Energy of Canada Limited (AECL) Underground Research Laboratory (URL) and at the Swedish Nuclear Fuel Waste Management Company (SKB) Aspo Hard Rock Laboratory (HRL). At both locations the excavations were tunnels in granitic material at approximately 420 m depth. However, the stress regime was more severe at the URL Mine-by tunnel site than the HRL ZEDEX tunnel. Different parts of the ZEDEX tunnel were created using different excavation techniques. AE and microseismic monitoring at the URL showed that events were most concentrated in the floor and roof of the tunnel, with less activity in the side walls. The side-wall activity was clustered primarily within 0.5 m of the tunnel wall. AE monitoring in the floor of the tunnel showed that small numbers of AE continued to occur in the notch region in the floor of the tunnel over two years after excavation was completed. This activity became more acute as the rock was heated, imposing thermally-induced stresses on the volume. Ultrasonic-velocity studies both in the floor and the wall of the tunnel showed that the velocity is strongly anisotropic with the slow direction perpendicular to the tunnel surface consistent with cracks parallel to the surface. The velocity increased with distance into the rock from the tunnel surface. In the floor, this effect was seen up to two metres from the tunnel surface. Most of the change occurred within the first 0.5 m from the tunnel perimeter. At the lower-stress HRL, most of the AE again occurred close to the tunnel surface. The occurrence of AE under relatively low stress conditions suggests that the regions experiencing AE activity were damaged during the excavation process, thereby reducing their strength. The section of tunnel excavated by tunnel boring machine had fewer events, clustered much closer to the tunnel surface, than the sections excavated using drill and blast extraction techniques. P-wave velocity changes of only about 0.1% were experienced due to the tunnel excavation for ray paths within zero to two metres from the tunnel surface indicating that crack damage was relatively low.