The main problem of tunnelling with use of TBM in highly dense urban areas is to assign the range of subsiding trough and the impact of tunnelling works on existing buildings and underground or road infrastructure. The paper presents the results of settlements calculations over twin tube metro tunnel using analytical, empirical methods. The tunnel external diameter is 6,5 m ; the overburden vary from 5 m to 8 m ; the distance between tunnel axis is 14 m. Because of quaternary soils and high water table level the TBM type EBP was chosen as the method of tunnel construction. At the length of 502 m of tunnel the monitoring system was carried out in 22 cross sections. The results of settlements monitoring were compared with the values of analytical calculations.

The impact of TBM EPB tunnelling was assessed with respect to the observed values of settlements as the results of extensive monitoring system of the subsoil and ground surface. The aim of the analysis using empirical methods was to determine the real scale of impact and to determine the formula for the asymmetric subsidence trough observed during the passage of two TBMs in quaternary cohesive soils. Based on field measurements, authors propose the polynomial formulation for the depth and shape of the asymmetric subsidence trough prediction over twin tube TBM tunnel.

One of major design problems associated with shallow tunnelling in urbanized areas is the prediction of ground displacements caused by the construction process. Advanced tunnelling techniques such as shield tunnelling using Earth Pressure Balance Tunnel Boring Machines (EPB-TBMs) allow for significant reductions of settlements observed at the ground surface in comparison to tunnelling methods used in the past. The predictions of these displacements are often based on semi-empirical methods and prior experience. In addition to relative simplicity of such methods, their robustness and decades of validation in many tunnelling projects make them attractive for practical use. The tunnelling-induced settlement trough at the ground surface can be described by inversed Gaussian distribution function. It requires only the assumption of two parameters, namely: expected volume loss (VL) and the distance to the point of inflection (iy ), which is dependent on the empirical trough width parameter (K) and the tunnelling depth (z0). The values of those parameters have a strongly empirical nature; they should be established based on comparable experience obtained from full scale tunnelling projects with similar technique and at similar ground conditions. The paper presents the problem of variability of those parameters and discusses the need for its assessment. As volume loss is strongly related to the tunnelling technique, the study focuses on EPB-TBM tunnelling as the most commonly implemented one in recent years. Variability of parameters observed for different ground conditions in different countries is summarized. Finally, preliminary assessment of variability of settlements observed in Warsaw region is presented.