Starting a large diameter TBM from surface

Abstract

In the Netherlands, when a bored tunnel has to be constructed, a launch and a reception shaft have to be excavated to guarantee sufficient cover. This is necessary to ensure face stability, prevent uplift and risk of blow out. Moreover, especially in the western part of the Netherlands, the first 10 to 20 meters of soil consists of the Holocene layer consisting of soft clays and peat. By starting the tunnelling process at sufficient depth, the tunnel will be driven through the Pleistocene and and the soil deformations will remain small. In 2009, a new method was introduced in Japan: the Ultra Rapid UnderPass (URUP) method. With this method the tunnel boring machine (TBM) starts tunnelling from ground surface level. Consequently there is no more need for a launch and reception shaft. This eliminates the need for expensive shafts which are time-consuming to build, which makes the URUP method interesting to apply in the Netherlands. However, applying this method will result in small covers at the beginning and the end of the tunnel lining. Therefore complications are to be expected. In this report, the feasibility of this method, applied for large diameter tunnels, is assessed by investigating the main complications associated with this method (tunnelling with marginal cover method) in the Netherlands. As a basis for the feasibility study a reference project, which is planned for the near future, is used. The reference project “Rijnlandroute” is situated in the west of the Netherlands (between Leiden and Voorschoten), where the first 11 meters consist of Holocene soft soil. For this thesis only the eastern start of the tunnel is considered, that is located in a polder. The groundwater level is located approximately one meter below surface level and the surrounding area is largely greenfield. Firstly, the soil deformations in the Holocene layer, when placing the TBM at surface level, are determined. It can be concluded that without ground improvement the settlements will become too big. It is therefore recommended to apply Mixed In Place columns, that have been applied successfully several times in the past for similar cases. When the TBM reaches the Pleistocene layer, ground improvement is no longer required and deformations remain small. Due to the high groundwater table an embankment is necessary in permanent situation to prevent uplift. Moreover, the tunnel invert has to be filled with a sand ballast as well. During construction (ballast is not constructed yet) the water level will be drawn down temporarily with a maximum of two meters. Due to the absence of cover at start of the tunnelling process, the force distribution in the lining differs from the conventional method. The axial loads are negligible, whereas bending moments still exist resulting in opening up of the joints, which might cause leakage. Therefore it is required to increase the capacity to withstand high bending moments when tunnelling with shallow covers. The required embankment will increase the normal force. However, this increase is too low to counteract the occurring bending moments in the lining. The Japanese box, which is a bolt type where the tension forces are transferred directly to the reinforcement, will be applied (permanently). This bolt type can withstand higher forces, though this bolt type will increase the construction costs. It can be concluded that despite these issues, tunnelling with marginal covers is feasible in the Netherlands. To gain more insight in the relative benefits of this method compared to the conventional method, both are assessed, considering the costs as well as the construction time. The comparison shows that the total construction costs only differ 1.8 % (the costs of the conventional method will be 1.8 % higher). However, the construction of the tunnel with the marginal cover method will result in a shorter construction time. The construction time of the conventional method is approximately four years whereas the construction time of the tunnel applying the tunnelling with marginal cover method only amounts approximately three years. Therefore, it will be efficient to apply this method in the Netherlands. However, despite the feasibility and efficiency of this method in the Netherlands, it cannot yet be concluded if tunnelling with marginal cover will be recommended in future. Further research is still required to conclude if the marginal cover method will be successfully. With this thesis it can only be concluded that the tunnelling with marginal cover method, applied for large diameter tunnels, is feasible in the Netherlands.