The role of overburden stress on the seismic vulnerability of deep tunnels G. Andreotti *1 and C.G. Lai 1,2 1 EUCENTRE, Via Ferrata, 1 – 27100 Pavia, Italy 2 University of Pavia, Dept. of Civil Engineering and Architecture, Via Ferrata, 1 – 27100 Pavia, Italy * Corresponding Author ABSTRACT. Generally speaking, underground structures have shown to be less vulnerable to earthquakes than above-ground facilities. This occurred even during severe events and the reduced vulnerability seems to increase with the increase of the overburden stress. Focus- ing on tunnels, this correlation has been shown by Sharma and Judd (1991) who studied empirically, the dependence between the level of expected damage and the construction’s depth. Their work, based on a database of 191 cases, showed that severe damage is restricted to depths less than 50 m while no appreciable damage is expected below 300 m depth. This trend does not appear evident in deep tunnels sur- rounded by weak rock. In the aftermaths of the 2008 Wenchuan earthquake in China, Li (2012) noted that whereas the damage suffered by tunnels in hard rocks was consistent with the findings of Sharma and Judd (1991), in weak rock collapses of tunnels were observed at depths as large as 500 m. For the same case study, slight damages were observed at 300 m depth. To clarify the role played by overburden stress onto the seismic vulnerability of deep tunnels in weak rock, this paper presents the results of a parametric study carried out using ad- vanced, fully-nonlinear, incremental dynamic analyses. The same model of the tunnel has been subjected to different levels of seismic in- tensity by varying parametrically the overburden stress and the properties of the surrounding ground. 1 INTRODUCTION Tunnels, being confined by the surrounding rock or soil, have long been assumed to have low vulnerabil- ity to seismic damage. Thus for a very long time, the seismic vulnerability of tunnels has not received the same attention of above-ground structures (Xiaoqing et al., 2008). The state of the art on the study of seis- mic response of tunnels appears delayed with respect to above-ground constructions, above all in the de- velopment of standard codes and software easily usa- ble to carry-out non-linear dynamic analysis. To date, the seismic analysis in tunnel design is not a common practice (Amberg and Russo, 2001) because tunnels are generally considered to be structures with earth- quake resistance (Towhata, 2008). Nevertheless this traditional point of view has been questioned by several strong earthquakes in the late twentieth century. The 1995 Kobe Earthquake in Ja- pan damaged heavily 20% of about 100 mountain tunnels near the epicentral area (Yashiro and Kojima 2007). During the 1999 Chi-Chi Earthquake, in cen- tral Taiwan, 26% of about 50 tunnels located within 25 km in the proximity of the seismic source were severely damaged while 22% were only moderately damaged (Wang et al. 2001). The 2004 Niigataken-Chuetsu Earthquake, in Ja- pan, affected about 50 mountain tunnels, including those that suffered minor damages, among over 100 that existed in the damaged region. About 25 of the affected tunnels suffered damages that required rein- forcement or repair (Shimizu et al., 2007). More recently, the 2008 Wenchuan earthquake in China severely damaged many tunnels located near the epicentral area (Li, 2012). This list is obviously not exhaustive because many other cases exist all over the world.