The city council of Amsterdam decided in November 1996 to permit construction of a new underground metro line, the Noord/Zuid-Lijn (north-south line). The two phases when complete will extend from the north of the greater Amsterdam region to Schiphol international airport.
The decision, finally ratified in 1997, was controversial as the project in its central section would involve bored tunnelling through the heart of historic Amsterdam. Nearly all the 1,500 buildings in the area affected by tunnel boring are of historical significance and of brickwork construction. Most are multistorey houses sensitive to settling; some already show signs of past damage. As well as these prominent buildings and structures, the section also passes below a number of canal bridges and quay walls. There is also a two-track tram line in the streets over the section.
Under the tunnel boring section are more than 20,000 pile foundations from the buildings and structures. There are also underground structures located directly in the axis of the line over the tunnels. The depth of the foundations varies and is based on the depth of the limiting sand stratum: (older) wooden piles reach an average depth of 12 – 15m while concrete piles (from the 1960s) usually have their foundations on a second sand stratum at a depth of 18m – 20 m. In both cases, the transfer of force from the piles via the point of resistance is decisive for settlement effects. The distance between the wooden piles in the foundation grid is approximately 1m. Owing to building remodelling/modernisation and renovation, it cannot be ruled out that unknown foundation remnants, or the like, are concealed underground.
Prevailing opinion was that construction of such a tunnel posed an unnecessary risk for these buildings and structures. Cut and cover, with its continuous annoyance to the public and damage to tourism over an extended period was not considered; many businesses and store owners feared a loss of revenue. Potential renovation of damaged buildings was rejected on the same grounds.
Despite public doubts, the need to enhance the city’s infrastructure allowed political will for the Noord/Zuid-Lijn to prevail and it was finally ratified by referendum in June, 1997. A positive outcome was achieved mainly through the conviction that advanced tunnel construction technology could reduce risks to an acceptable level.
Seldom has this technology been employed in such a challenging environment, which meant that potential consequences had to be investigated very early. It quickly became apparent that the heightened demands placed on process quality could only be met, and risks be anticipated and controlled, through the use of an exceptional settlement-oriented shield drive system. This marked the birth of a process control system which developed into the concept of the Integrated Boring Control system (IBCS).
The project
Initial execution will begin in 2001 comprising the above-ground section in the northern section of Amsterdam, the immersed tunnel below the IJ and Central Station, bored tunnel under the entire historic part of the city and an above-ground segment to the World Trade Center in the south of Amsterdam. When this 9 km long section is completed in 2008, it is anticipated that the new line will transport 200,000 passengers/day.
The bored tunnel segment encompasses the section south of the Centraalstation (central station) to the area near Europaplein. Length underground is approximately 3.8 km, including three deep stations: Rokin, Vijzelgracht and Ceintuurbaan. Of this total length, around 3,150m must be constructed by TBM technology in two individual single-track tubes with an inner diameter of 5.82m and a borehole diameter of approximately 7m.
Geology and hydrology
The global soil structure in the region of Amsterdam comprises quartenary alluvium and Pleistocene Period layers. The top layer is made up primarily of fill, Wadafzetting (banks of sand deposit) and Basisveen (peat).
Below this top layer is a first sand layer, then an Intermediate Alleröd layer and then the second sand layer followed by Eem Clay (clay bed). Tunnel driving will take place almost completely in the last three layers.
A local exception applies to the Ferdinand Bolstraat, where Basisveen and sand deposits containing clay are anticipated. Of special significance for overall advancement of the tunnel, despite transformation and transgression to other layers, is the Eem Clay, which, on account of its low modulus of elasticity, reacts immediately with significant deformation to changes in stress. The groundwater table is 1m below ground surface. Individual tunnels are subjected to a water pressure of around 3.5 bar, with maximum support pressures of up to 5 bar required.
Alignment
The section below historic Amsterdam must often follow narrow streets. Tunnels mostly run adjacent to one another at a regular distance of one half diameter, although between Vijzelgracht and Europaplein they are routed one over the other through the Ceintuurbaan station. Curves, down to a smallest radius of 190m, also vary.
In sections with falling grades, with slopes of up to approximately 4%, the tunnels will reach depths of around 35m, whereas the minimum cover below the body of water Rokin is approximately 15m.
Process analysis and new methods
In the course of process analyses and development work it also became evident that the current existing shield drive approach would have to be supplemented with process equipment which permits discrete implementation of all the determinative process control tasks.
The Noord/Zuid-Lijn project team has compiled diverse, helpful innovations into one integral concept which aims to stimulate innovation from contractors.
Early settlement risk assessment (SRA) studies for the project revealed that state-of-the-art tunnel boring methods, with a 1-3% volume loss, would create an unjustifiable risk of settlement. Additional and expensive mitigating measures would be required for almost the whole length of the tunnel. To avoid this meant a search for alternatives and thorough re-investigation of the tunnel boring process. Parallel to this, the first shield drive project in Holland, the Heinenoord tunnel, was used as a large-scale testing ground for verification of the theoretical model calculations that reflect the risk of (pile) settlement.
Analysis of the potential for innovation of the existing boring techniques was begun at an early stage in the planning and some initial results were presented during the STUVA conference in Berlin in December, 1997 à.
The innovation options up to that time were compiled into an integrated concept, the result being the compact version of the variable (vario) shield system.
The concept allows for various mechanical components or solutions for dealing with and mastering the tasks associated with the specifics of the boring system. It is a variable and adaptable shield system which meets the first prerequisite for fulfilling the highest quality demands during tunnel boring. The cutting wheel and working chamber accommodate slurry as well as work in EPB mode, while the muck transport system is directly connected to the working chamber and therefore can be changed from slurry pipe to screw conveyor at any position in the track, under atmospheric conditions. The rigid length of the machine is very compact (only 50% of the length of “regular” TBMs of this diameter) and allows for very small curves without soil deformations. (See lectures at the STUVA and ITA conferences à, á.)
However high performance a shield system is, it remains just a tool. The final quality of the tunnel product is determined by how the tool is used. Against this background, high-level process quality was given special consideration during planning and development of an integrated process control system was begun.
IBCS philosophy
The diverse characteristics for tunnel boring for the Noord/Zuid-Lijn demand sensitive operation for the tunnel boring system with continuous control of interaction with the soil and further interaction with the pile foundations, including any resultant damage to buildings.
To better master these complex tasks and execute process decisions in a settlement-oriented manner, tunnel boring personnel must be supported appropriately. Support is based on two pillars: provision of up-to-date information about relevant process effects, and automation of determinative process sequences.
Due to the complex correlations between the various individual processes an integral process management system must be employed. Modern computer systems drawing on a wealth of tunnel construction experience can perform the extensive data processing operations necessary. Adapted to the specific requirements of tunnel boring for the Noord/Zuid-Lijn project, this is the integrated boring control system (IBCS).
Many open-and closed-loop control systems are already in use in current tunnel boring systems. The distinctive feature in the IBCS concept, is the on-line integration of settlement information into the control parameters: tunnelling is not controlled solely on tunnel and machine data. Moreover, settlement caused by tunnel boring, is used simultaneously with these as a control criterion. To overcome the delay in real process and real settlement the actual boring process is also gone through in a virtual environment and provides real-time predictions about settlement which are continuously verified using check measurements in the surroundings (ground/buildings/structures).
The IBCS requires a number of components for model simulation of the boring process with changing input parameters, for virtual execution of the process, for predicting the effects of the process, for comparison of the actual effects of the process, for evaluation of the comparison results and for adjusting the actuators of the boring system. These components are categorized into three functional units:
1. Effect measurement appears relatively independent in the IBCS; the crucial point is that the real effects which are passed on in the system provide unique information and that they can be allocated to a certain time. Using automatic monitoring linked to the IBCS on-line, continuous and simultaneous data acquisition for the various individual processes is provided during tunnel boring. All information is fed into a geographic information system (GIS), which is in the final stages of development for the Noord / Zuid-Lijn project â. Using this system, accurate visualisation of the results can be realised automatically for rapid analysis and interpretation.
2. Evaluation and Prediction analyses the effect data which has been measured, evaluates this data and stores it for the predictive calculation and subsequent comparisons. Using a 4-D FEM computer model in which the determinative process parameters for the TBM are modelled, the effects of the boring process are simulated by means of a virtual boring process.
For this, the bore front support system and the annular void support, among other things, are modelled as active pressure-controlled liquids which pass their reaction forces (established through calculation) onto the shield and soil continuum. This permits predictions to be made concerning subsoil deformation and surface settlement. In this process, the 4-D computer model also serves to simultaneously couple the predictions with the measured data for the effects of boring and the process settings, resulting in a closed, real-time information circuit for the integral system. Adjustment of the process parameters to the actual effect data permits the process sequence to be successively optimised with respect to generation of settlement.
3. Decision making and Adjustment. In the initial development stage, this unit is designed as a purely information and visualisation system for boring process data. Decision making and adjustment must continue to be carried out by humans in an expert team.
The TBM measured data are evaluated for this purpose and the relevant boring process variables filtered out. Real-time visual display of all this data at the control panel, along with the determinative real and virtual effects produced by the boring process, forms the basis for decision making by the process control team.
In the semi-manual operating mode, the operating personnel must think in terms of process variables rather than in terms of actuators on the control panel (ie for example the pressure in the tail void rather than the pressure at the grout pump).
It means they have to direct their decisions, based on experience, by the effect on the boring process. Experience gained with TBM data is, hence, transferred to the boring process parameters, to which setting of manipulated variable changes is then also oriented.
It is planned to replace this decision making and setting process with calculation algorithms for an automatic knowledge based process control system later.
In its current draft phase, the virtual model is being calibrated and validated by comparison with results for the test pile field at the Heinenoord tunnel. Using the variable compact shield as a calculation model in 4-D computer simulations, the validated prediction unit is, in the current planning phase, already providing information about the virtual effects of the newly developed TBM concept and about ground deformations of the modelled region, as the input for SRA studies.
Adapted Mechanical Systems
In order to implement settlement-oriented control of essential individual processes, these processes require a closed information circuit. In addition to the function “Control”, which can be used as a functional unit in the IBCS, the other functions “Measure” and “Adjust” are also needed for completing the circuit. These two functions must in essence be performed by the hardware, ie the mechanical systems. Additional innovations for current tunnel boring methods need looking at.
Here, only the example of the annular void support process is dealt with. Based on the concept of the variable compact shield, a newly created option for “Measuring” and “Adjustment” is illustrated.
For settlement minimisation, it is important to actively influence the interaction between boring and the ground. This means that process variables must be recordable and/or measured and also be controlled and/or adjusted.
However, nowadays there is no possibility for reliable acquisition of the control variable “Support pressure in the annualr void”. And so an innovative approach is needed: the principle of active shield support in which ground pressure around the shield body is counter balanced using an intermediate pressure medium between the shield jacket and the ground.
Rather than a constant cylindrical form, the shield body has two regions with designed differences in diameter.
The front region, for passive support, defines the hollow area in the ground by its larger diameter shield section and simultaneously generates an annular void around the rest of the rear tapered shield body, that is simultaneously filled with a liquid support medium for active support.
This intermediate support medium matches the locally changing form of the hollow area around the shield body in this process without any loss of volume and transfers the support reactions from the ground surface to the shield without triggering any soil deformation.
The pressure in this medium forms the process variable that can be acquired on the circumference of the shield using a number of pressure sensors. Supply of the medium can also be actively regulated using these variables. Adjustment of the pressure level to the prevailing ground and water pressure ensures an optimal counterbalanced state.
A practical design solution is an articulated attached shield tail with a special tapered-shaped longitudinal section. With the tapered-shaped leg, the shield tail closes off the annular void at the face, while functioning as a temporary gliding inner shell . Such a shield tail inner shell does not produce direct, rigid support of the immediate surrounding area. Instead, the annular void compression medium also functions as the intermediate support medium while, at the same time, retaining the process advantages for “Measuring” and “Adjustment”.
The axial cylinder attachment of the shield tail, combined with an oil-hydraulic storage circuit, uncouples the pressure control process from tunnel advancement; the result is self-regulating compensation of the volumetric flows and, hence, elastic pressure stabilisation such as that used in extrusion concrete systems.
The shield tail inner shell offers other operating advantages which also have a positive, minimising effect on settlement. (See STUVA conference à on the variable compact shield). Despite all the theorectical confidence of the planning team there nevertheless remained a healthy degree of doubt concerning the realisation of the high quality demands of the project. Following their motto “design by testing”, they initiated a “full scale” tail void injection test in which the grouting concept will be tested. In this 28t test facility the thickness of the tail void is on a 1:1 scale although the diameter is on a 1:5 scale.
Conclusion
Development of the integrated boring control system continues. Achievements so far confirm its general feasibility and also provide a glimpse of the anticipated benefits, even when the IBCS does not function as a control system, but only as an information system. It provides details about the various interactions during tunnel boring and, in any event, furnishes the operating personnel with insights into the effects of the boring process and thus presents new, important decision making criteria for process control.
Complete functionality as a control system for the final stage of development can only be implemented when the mechanical systems have been suitably adapted. As the example presented here illustrates, ways and means must be found to integrate all of the process-relevant parameters into the closed information circuit of the control system. The innovation analysis of the Noord/Zuid-Lijn Consultants provides an initial, interesting impetus for further detailed studies on this topic.
Particularly beneficial in the development of the IBCS, or better, for its implementation, is the fact that components for both the IBCS and for the mechanical systems can be used independently as modular applications. The mechanical systems innovation required for the IBCS provide the operating personnel in semi-manual operation with improved prerequisites for process control. In no phase of operation is it necessary to make an “all or nothing” decision; there is always the option of utilising intermediate developmental stages with a reduced innovation risk to good effect. Likewise, the loss of individual components used at a later stage does not result in a total loss of the tunnel boring system. Process quality is reduced though it remains higher than without the IBCS.
Development of integrated boring control also forms a basis for interesting innovations on less demanding projects
Related Files
The tunnel alignment (vertical)
Functional units of the IBCS
Process structure of the integrated boring control system – IBCS