Soil conditioning and lubrication are commonly employed during tunnel construction, with applications for earth pressure balance (EPB) tunnelling machines, and for pipe jacking construction methods. These techniques involve the introduction of chemical conditioning agents to modify the soil properties during excavation, and when applied effectively can lead to significant improvements in tunnel machine performance and the overall construction process. For tunnel excavations with EPB machines, conditioning agents ideally change the excavated soil into a soft, plastic paste. This allows improved control of the excavation process and head chamber pressure, reduces torque and machine wear, and reduces clogging of the machine by sticky soils. Lubricant fluids are used during pipe jacking to reduce interactions between the pipes and soil, resulting in lower jacking forces and allowing longer drives.
A wide range of soil conditioning agents for EPB machines are available, the most common being foams and polymers, but little guidance exists concerning the selection and application of conditioning agents for tunnelling in different soils. The various types of conditioning agents interact with different soils in different ways, and can be used separately or in combinations at various concentrations and injection ratios. As the properties of a conditioned soil depend on the specific treatment, the number of different agents and variables in their application make the optimum treatment for a soil difficult to define. Similarly, the required properties of pipe jacking lubricant fluids depend on the ground conditions, and the range of materials and chemical additives available make determination of an optimum lubricant fluid formulation a difficult task.
A research project at Cambridge University is investigating soil conditioning for EPB machines and lubrication for pipe jacking, focussing on their applications for tunnelling in clay soils. A number of issues are being investigated, including testing of soil conditioning foam agents, conditioning of London Clay with polymers and foams, model EPB machine screw conveyor tests with conditioned clay soils, and model pipe jacking tests with various lubricant fluids in clay soils. Field monitoring of tunnel machine performance and soil conditioning applications is also being performed on the CTRL project in London, UK.
Foam agent index tests
There are many soil conditioning foam agents available, marketed for applications in different ground conditions and with recommendations for typical ranges of concentration, foam expansion ratio and injection ratio. A number of foam agents have been tested by measuring index properties of the foams, produced with a laboratory foam generator. The foam expansion ratio (FER) and liquid drainage time of foam samples were measured to compare the properties of foams produced from different agents. The FER is the ratio of air to liquid volume in the foam, measured through the mass of liquid in a foam sample of known volume, and the drainage time measures the time for the liquid to drain from the foam, indicating the stability of the foam.
Five foam agents were tested, each at two concentrations based on the supplier’s recommended ranges, shown in Table 1. These foam agents are based on anionic surfactants or glycols, with some containing polymer additives. To compare the foam properties through the index tests, all samples were produced under the same conditions with the foam generator. Figure 1 shows the FER measured at atmospheric pressure for each foam agent concentration, and the measured T50 values representing the times for drainage of 50% of the foam liquid.
The FER of the foam samples varied considerably between the different agents. The FER increases with concentration for each agent, but they produce foams with different ranges of FER over different concentration ranges. These differences are presumably a result of the different chemical compositions of these foam agents. Comparison of these results with the supplier recommendations in Table 1 shows that the expansion ratios measured for each agent are within the stated ranges. For some agents a high FER could be achieved at low concentrations, and to achieve the lower recommended FER values, concentrations below the suggested values could be used. The FER can also be varied through the foam generation parameters, with the air and liquid flow rates and pressures all having an influence, as well as the foam generator design details. While the results shown here are specific to the foam generator used, they indicate that for a given foam agent a range of FER can be achieved by varying a number of factors in the foam production.
The foam liquid drainage times in Figure 1 show that Foams C, D and E all had similar drainage times, with 50% of the liquid drained from the foam bubbles in less than 10 minutes. For these foams the drainage times were not significantly affected by the foam agent concentration or the expansion ratio. Foam D contains a polymer additive, which is supposed to increase the foam stability, but this effect was not apparent through the liquid drainage time. Foams A and B were significantly more stable than the others tested, with 50% drainage of Foam A taking up to one hour. For these foam agents, the drainage times increased with both concentration and expansion ratio. Again, the different properties of these two foam agents are likely to be due to their specific chemical compositions.
London Clay conditioning index tests
During tunnelling with EPB machines, conditioning agents are injected as the soil is excavated and passing through the machine, ideally mixing with the soil to produce a soft, plastic paste with undrained strength in the range 10kPa-25kPa(1). The optimum treatment to create an effectively conditioned soil depends on many factors and varies widely for different soils, and so is usually determined by trial and error. Index tests have been performed to measure properties of London Clay cutting samples conditioned with various treatments.
The clay cuttings were obtained from a shaft excavation for the CTRL project. They were sieved so that the maximum size was 25mm (for the purpose of these laboratory tests) and mixed, at natural moisture content, with polymer solutions and foams at various injection ratios to represent the conditioned soil cuttings created as the machine excavates the soil. Shear vane and large-scale fall cone tests were performed to measure the undrained shear strength of the compacted, conditioned clay samples.
A series of index tests were performed on samples prepared with three polymer conditioning agents, each at two solution concentrations and a range of polymer injection ratios. When mixed with the soil, the polymer solutions bind the clay cuttings together to form a uniform, soft paste. Figure 2 shows the vane shear strength measured for samples with polymer injection ratios from 10%-60%, and for samples prepared with water for comparison. The polymer conditioned samples had lower strengths than those conditioned with water at these injection ratios, and the polymer solution concentration did not significantly affect the strength. Polymers A and B are both PHPA (partially-hydrolysed polyacrylamide) polymers from different suppliers, and result in similar conditioned sample strengths, while Polymer C is a different chemical and resulted in slightly higher strength samples.
To investigate conditioning of London Clay with foams, index tests were performed with two foam agents over a range of injection ratios and at different expansion ratios. Figure 3 shows the vane shear strength measured for samples prepared with Foams C and D, with the foam injection ratios increased to reduce the strength to suitably low values. These foams were found to perform poorly when mixed with the clay cuttings, as the foam liquid was absorbed by the soil causing rapid breakdown of the foam bubbles. The foam liquid softened the clay cuttings, but very high injection ratios, several times greater than the volume of the soil, were required to create a soft paste. These foams had little effect on the London Clay when injected at the volume ratios (FIR) recommended by the supplier shown in Table 1. Foam D contains a polymer additive in the liquid phase, but this agent did not show improved performance compared to Foam C, as the amount of polymer added to the soil through the foam liquid is small. Samples were prepared with Foam D at two expansion ratios, so different amounts of foam liquid were mixed with the clay cuttings. The samples prepared with FER=10 had similar strengths at approximately half the foam injection ratio of the samples prepared with FER=19. These results indicate that similar strengths are achieved when equal amounts of foam are added to the soil through the combination of the foam expansion ratio and injection ratio.
The results of these index tests indicate that the most successful conditioning treatment for London Clay was with polymer solutions at injection ratios in the range 20%-40%, creating soft pastes with strengths of approximately 20kPa-5kPa. The collapse of the foams when mixed with the London Clay cuttings resulted in only the foam liquid conditioning the soil, requiring very high injection ratios to be effective and eliminating other benefits of foam conditioning resulting from the presence of the air bubbles throughout the soil. In practice foams are widely and successfully used as soil conditioning agents for many types of soils, but these tests indicate that they are not the most appropriate conditioning agents for stiff, high plasticity clays.
Model EPB screw conveyor tests
The screw conveyor of an EPB machine has a critical role in the excavation process, controlling the volume of soil discharged from the machine and the head chamber pressure supporting the tunnel face. Effective conditioning of the soil is important to achieve controlled operation of the screw conveyor by forming a plug to balance the earth and water pressures in the head chamber, and to minimise torque requirements and wear.
To investigate the mechanics of operation of the screw conveyor and the effects of soil conditioning treatments on the performance, a model EPB machine screw conveyor system has been developed at Cambridge University. The model screw conveyor is approximately 1:10 scale of a typical EPB machine conveyor, with a length of 1m and diameter of 0.1m. The horizontal conveyor connects to a pressurised container filled with the soil sample, from which the screw carries the soil along the conveyor to a discharge outlet. The conveyor casing tube is instrumented at four sections, each with two load cells measuring the total normal stress and components of the soil-casing interface shear stress, and a pressure transducer measuring the pore water pressure in the soil. The screw drive shaft is instrumented for measurement of the torque during operation. Test samples are prepared from consolidated pure kaolin clay, or from compacted pre-conditioned natural soils. Tests can be performed with variable pressures applied to the soil, at a range of screw speeds, with different discharge outlet conditions, and with model screws of different geometry.
During operation of the screw conveyor, the pressure applied to the sample is dissipated through shearing of the soil against the casing and screw surfaces as the soil is carried along the conveyor. When the screw conveyor is discharging soil with the screw rotating at a constant speed and with a constant pressure applied to the sample, a steady state is reached where the pressures measured at each instrumented section are stable over time. Figure 4 shows measurements from a test with a sample of compacted London Clay cuttings, conditioned with a polymer solution at 50% injection ratio to give an undrained shear strength of 5kPa. A constant pressure of 200kPa was applied to the sample, with the screw rotating at 15rpm. The total pressure dissipates linearly along the conveyor, from the applied pressure of 200kPa (representing the pressure in the TBM head chamber), to approximately atmospheric pressure at the discharge outlet. Also shown in Figure 4 are the pore water pressures measured at each section, which also reduce linearly along the conveyor in response to the total pressure changes, resulting in the normal effective stress remaining approximately constant along the conveyor. This results in an approximately constant shear stress measured along the conveyor. The torque to rotate the screw also remains constant during stable operation of the conveyor at a given speed. This general mechanical behaviour has been observed for a wide range of tests performed with clay soils, and accounts for the linear dissipation of pressure observed along the screw conveyor.
In addition to the laboratory research on the conditioning of clay soils and model screw conveyor tests, field monitoring of the performance of EPB machines and the conditioning treatments used for different soils is underway on CTRL. This research is providing data to compare the laboratory tests with the performance of the tunnelling machines and soil conditioning treatments in the field.
Pipe jacking lubrication research
Conditioning agents also find application in pipe jacking. There, the intended benefit is to decrease jacking forces by injecting a lubricant into an overcut around the pipes during their installation so as to prevent the build-up of excessive radial stresses on the pipes. In stiff clays, stress relief and swelling occurring around the cavity subsequently to the excavation can result in large pressures on the pipes and, in turn, high jacking forces. The injection of a water-based lubricant such as bentonite around the pipes provides water that can readily be absorbed by the surrounding clay. This measure to reduce jacking forces can prove counterproductive. Recent approaches to counteracting this process have suggested the use of chemically enhanced bentonite slurriesà. Chemical agents acting as swelling inhibitors, such as polymers, salts and formates, can reduce the rate of water intake into the ground and thereby delay the swelling process and the associated build-up of jacking forces.
Although some successful applications of such chemicals in pipe jacking have been reported, little evidence exists as to the net effect of these chemicals on jacking forces. The mechanisms relevant to the interaction between the ground, the pipes and the chemicals are not yet sufficiently understood to allow guidance for the use of these conditioning agents in different ground conditions.
The effectiveness of chemical additives in reducing the magnitude of jacking forces needs to be investigated. Understanding of the properties of the ground treated with different soil conditioning agents is required to allow guidance for the selection of the most appropriate lubricant composition for each ground condition. Furthermore, the interactions between the size of the overcut, the type of lubricant and the injection pressure of the lubricant around the pipes are not yet fully understood. Increasing the size of the overcut can result in unacceptable ground movements and pressurising the lubricant with the aim of supporting the ground can actually accelerate the swelling process, hence proving counterproductive. It is the aim of this research into pipe jacking lubrication to investigate these different interactions.
The laboratory-scale pipe jack was devised for this purpose. This apparatus simulates the pipe jacking process with a certain degree of idealisation. It allows the insertion of an instrumented model pipe into a clay model of controlled stress-history. The model pipe is equipped with an excavation tool similar to that used on a self-boring pressuremeter and includes a slightly oversized cutting shoe (simulating a shield) that leaves a deliberate gap around the model pipe, into which lubricant may be injected in a similar way as in pipe jacking. As shown in Figure 5, the apparatus is instrumented for measurement of the radial stresses and the pore pressures on the pipe, pore pressures at various locations in the ground, as well as the axial force required to jack the model pipe during its installation and at subsequent stages leading to long-term equilibrium. Finite element analyses and field monitoring of pipe jacks in London Clay are being carried out concurrently in order to test the validity of the results from the experimental work.
Related Files
Fig 1 – Foam expansion ratio and liquid drainage times of various soil conditioning foam agents (solution concentrations in brackets)
Fig 2 – Shear strength of London Clay samples conditioned with various polymer solutions (solution concentration in brackets)
Fig 5 – Schematic cross-section through the model pipe
Fig 3 – Shear strength of London Clay samples conditioned with foams
Fig 4 – Pressure changes measured along screw conveyor during a test with conditioned London Clay
