Movements in the Piccadilly Line tunnels due to the Heathrow Express construction

Abstract

This paper presents the results of analyses of monitoring records taken from inside one of the existing Piccadilly Line tunnels during the construction of three station tunnels at the Heathrow Express Central Terminal Area. The monitoring data allow a detailed picture of the way in which existing segmental concrete-lined tunnels behave as a result of tunnelling works below. Settlement curves at critical construction stages are presented. It is shown that the computed ‘volume loss’ varied between 1·3% and 2·5%. Long-term settlement records indicate that the maximum settlement increased by 27% for a period three years after tunnel completion. This result is compared with earlier findings from the Heathrow trial tunnel. Asymmetry of the settlement troughs caused by the outer tunnels, constructed after the central concourse tunnel, is discussed, and asymmetry parameters are defined from a number of case histories. The relationship of asymmetry parameters to pillar width between tunnels is presented and discussed. This suggests that interaction between the adjoining tunnels becomes negligible when the separation is six to seven diameters. Rotation in the existing tunnels at various construction stages is described, and the progressive rotation of the existing tunnel immediately above the concourse is shown. A tunnel longitudinal section showing ground movement contours predicted using a semi-empirical method illustrates the development of rotation. Distortion patterns along the existing tunnel, obtained from tape extensometer readings taken after completion of new tunnelling, are illustrated. These results together represent the first comprehensive study of tunnel and lining response to adjacent tunnelling published in the literature. The patterns of settlement, rotation and distortion, when considered together, provide a graphic description of a complex soil–structure interaction response to stress changes in the ground. The data can be used to predict responses in other soils and for other geometries by extrapolation.