The Westerschelde tunnel will be the first traffic-tunnel to link the Dutch provinces Zeeland, Zuid-Beveland and Zeeuws-Vlaanderen. The 6.6km connection comprises two separate 10.10m id tunnels driven by Herrenknecht slurry shield TBM with an outer diameter of 11.33m. A substantial part of the tunnel runs through the Boom clay (Boomse Klei), an over compacted, mostly stiff, tertiary clay. These difficult ground conditions and water-pressures up to 6.5bar had to be taken into consideration during the design of the TBM (1).
The paper first discusses the “stickiness” specifications from the point of critical operational conditions. It also describes technical innovation and development of the two TBMs. With this information, characteristic operational parameters were set for early identification of problems. Finally, a comprehensive data-acquisition system was set up to monitor the tunnel-lining process.
Critical geotechnical conditions
The geology of the site is characterised by alternating sand and clay-formations. Under these sand layers is the Boom clay layer, known to be highly adhesive. During projects with similar soil conditions this stickiness has caused difficulties and resulted in a poor progress of the TBM.
Here about one-third of the alignment runs through sand; one-third through the Boom clay and one-third through sand-clay.
Definition of risks
Three elements characterise stickiness in the working condition: adhesion, cohesion and dispersion.
Adhesion, which is the degree to which soil will stick to the steel of the shield, is responsible for problems like clogging of cutter units in the mud pressure chamber, and closure of the cutting wheel openings.
Cohesion describes the tendency of the soil to coagulate. Common results are the blockage of the working chamber or sometimes of the transport facilities.
Using a slurry shield, hydraulic aspects are of fundamental importance. Therefore special attention should be given to the hydraulic properties of the bore fluid. These properties can be described through the rate of dispersion.
Until now there have been no standardised criteria to distinguish different rates of stickiness. In his paper Thewes (3) put forward the theory that the water rate and the critical conditions of the soil are major factors, regardless of the way the soil research has been carried out. The stiff and half stiff layers of the Boom clay are expected to cause problems like adhesion.
Machine design
The results of detailed laboratory studies contributed to improvements of the cutting head, mud-pressure chamber and suction area. The bore fluid has also been refined by data from existing hydroshields.
The Westerschelde machines have alternative bentonite supply route other than through the bars of the rock crusher. These are:
Additionally, two agitators are placed right in the suction area to prevent a build-up of material.
Definition of critical working conditions
In scenario 1, the speed of the TBM above a critical limit results in a build-up of soil in the mud pressure chamber (phase 2). The bentonite supply in the suction area only circulates within the chamber, instead of transporting the material. If the speed increases to phase 3, the soil in the mud pressure chamber thickens. In the worst case (phase 4), the whole suction area will be blocked. It is then necessary to clean the machine under high air-pressure before it can be used again.
In the Westerschelde project, cleaning is only possible by divers entering the mud chamber which is very good reason for wanting to make such cleaning an exception.
In the scenario 2, delay is caused by an increase of density. Suspension flow near the bulkhead is too small (2) and the large clay blocks require mechanical devices to move them in front of the cutter hole. A shortened holding time in the mud pressure chamber decreases the softening of the clay.
This increase in clutching of the clay can lead to a complete blockage of the suction area in phase 3.
So when considering actual operation, it is not only cohesion and adhesion that are important, but the fluid mechanical characteristics in the mud pressure chamber.
The speed of the bore fluid leaving the slurry pipe is about 1.5m/s to 3.7m/s. The speed of the bentonite-suspension near the dividing wall openings depends solely on the conditions in the mud pressure chamber.
But even with a 100% supply from the diving wall, this gives only a maximum velocity of 0.14m/s near the suction area. The drag forces that can be executed on large clay lumps are too little to guarantee transport of material without delay. Agitators are placed to assure a good suction flow.
Data acquisition
The analysis of the causes and identification of the difficulties around thickening is complex. Therefore, on the Westerschelde project, a major data collection exercise has been carried out. The aim is to recognise critical conditions at an early stage. A monitoring system will establish a relationship between the mud pressure and the progress of the TBM. Other systems will monitor the material flow and the density of the bore fluid in comparison to the soil.
From these parameters, tolerable deviations from the mean can be calculated. When this is exceeded, measures have to be taken immediately.
The overall monitoring system is based on elementary principles. For instance, if over a set time period, the quantity of material entering the mud pressure chamber is equal to the amount leaving, then stickiness of the soil is impossible.
One current research programme is aimed at improving measurement methods. Analyses of equipment such as the radiometric density-meter is taking place at the Lehrstuhl für Bauverfahrenstechnik, Tunnelbau und Baubetrieb (Department of construction techniques, tunnelling and construction management) at the Ruhr University in Bochum in cooperation with the Herrenknecht.
Measurement by the second monitoring system offers one possible means of understanding stickiness in the area of the submersed wall hole.
Stickiness of soil causes a resistance in a system of two communicating pipes. So if stickiness occurs in the submerged wall this results in pressure deviation between the mud pressure chamber and the working chamber. It is also possible to analyse the pressure differences between the mud pressure chamber and the suck pressure of the pumps further on.
The secondary monitoring system is based on a progressive measuring of the bore fluid pressure in the mud pressure chamber. From the bore fluid pressure figure it is possible to calculate its density and compare it with the standard.
To judge the risks of sticking in the mud pressure chamber another option is to measure the earth pressure in the rotating cutter wheel. From the minima and maxima of the sinusoidal pressure curve, it is possible to determine the density of the ground and of the bore fluid. Disturbances in the sinus reveal local variations of density and therefore of stickiness.
Data evaluation
With specialist evaluation of the data of the monitoring system, mean values of production and tolerable deviations are determined.
Preventative measures are mostly focused on determining the critical aspects of progress and the choices available in optimising the bentonite supply circuit.
The bentonite supply is automatically controlled. So when, for example, stickiness of the suction area is detected, the TBM driver can adjust the bentonite supply within the critical area, changing fluid properties to increase the flow of the soil.
Experiences until now
With some 3800m of the tunnel complete – some 1800m in the Boom clay – the experiences gained at Westerschelde are mostly positive. Supply of bentonite or water at the diving wall in the mud pressure chamber is 100%.
Tangential supply in combination with the agitators and the rotating stone crusher have proven to be very effective. The progress speed in Boom clay is about 25-35 mm/min and it is primarily restricted by the capacity of the separation facility. Disturbances in the mud pressure chamber and the rest of the cycle can be clearly defined and recognised. A critical look at the balance of mud and bentonite can also give good results.
Future perspective
Stickiness is primarily a failure in the steering of the TBM. The newly introduced method gives some new standards in diagnosing stickiness so preventative measures can be applied at an early stage. Because human error cannot be ruled out, it is useful to equip the machine with modern shields with an additional cleaning system.
In case of major coagulation in the mud pressure chamber, which is not controllable with the bore fluid, the TBM is shut down. Horizontal moving high-pressure pipes can be brought in from the shield into the mud pressure chamber under atmospheric pressure. A rotating water jet can disperse piles of mud with a diameter of 1.2m. The positioning of pipes has to be focused on critical zones and the cutter-wheel sector.
Related Files
Schematic diagram of the two stickiness problem scenarios
Injection points and stone crusher on the TBM
Geological cross section
Secondary monitoring system
Pressure chambers
Tunnel cross section
