Subsidence Due to Groundwater Leakage into Tunnels

Abstract

When rock tunnels are constructed under populated areas, additional geotechnical challenges that could often be met are ground subsidence due to groundwater leakage into the tunnels. The effect of such subsidence is a serious issue with regard to the infrastructures built around the tunnel areas. Subsidence analysis as adopted, today, for various transportation projects is based on classical consolidation theory. The increase in effective stress, which comes from groundwater-lowering, is assumed to induce an equivalent excess pore pressure that results in consolidation settlement. The settlements are then computed based on the classical theory of one-dimensional consolidation. However, in the particular type of problem, the total stresses in soil remains practically unchanged throughout the subsidence process. Therefore, excess pore pressure is not expected to be generated. In other words, the classical theory of primary consolidation alone under-predicts the subsidence. A real case example, the Hommelvik tunnel, from Norway is considered in this study to elaborate this. The paper presents subsidence data gathered during the past 20 years from representative locations at the tunnel site. The results show that the area has been subjected to larger subsidence than expected. This paper emphasizes the importance of prediction and instrumentations to understand resulting subsidence over a large area due to tunnel construction. The results obtained from the finite element (FE) analyses are discussed in light of these measured data and background assumptions. FE analysis using a soft soil creep model is seen to give a very promising framework to analyse such problems. This particular study calls for a better subsidence prediction procedure using a framework that also considers the effect of creep.