Behavior of over-cutting area injected by flyash surface-active agent mixture material

1. INTRODUCTION

Pipe jacking is a technique used to install underground pipelines through a borehole excavated by a shield-type boring machine, which operates hydraulically from a launching shaft. Slurry pipe jacking is firmly established as a special method for non-disruptive construction of underground pipelines. The advantages of this method have been widely recognized [1]. During the pushing processes, the mud slurry is injected into the face and over-cutting area surrounding the pipes. After completion of the pipeline, the mortar is injected into the over-cutting area in order to provide permanent stability of the surrounding soil.

During the jacking process, the jacking load is an important parameter in controlling the pipe wall thickness, location of intermediate jacking stations, selection of the jacking frame and fulfilling lubrication requirements [2]. Reduced jacking force further results in lowering axial loads placed on the pipes, which minimizes the risk of damage during installation.

Lubrication can effectively reduce the needed thrust if a discrete layer of the lubricant is maintained between the pipe and the excavated soil. Lower frictional force allows larger jacking lengths to be achieved. Reduction of frictional stress around the pipes is closely related to the efficiency of lubrication injection. The lubricant must be designed so that it can form a layer in the surrounding soil, be pressurized to overcome ground water pressure and stabilize the over-cutting area. Lubricant should fill the complete over-break void to minimize surface settlement.

However, it was clarified that for a commercial lubricant to be effective, it is suggested that ingredients such as sodium and potassium be eliminated. As a result, for a lubricant to maintain its functionality as a means to mitigate the overburden pressure of an over-cutting area, and to enable a reduction in thrust, it is necessary to develop a better quality lubricant which can overcome these problems as soon as possible.

Taking this into account, a lubricating material comprised of a mixture of the surface-active agent with the flyash was developed [3]. In this paper, in order to examine the characteristics and behavior of the flyash surface-active agent mixture developed as a lubricating material, a fundamental investigation is performed and these results are discussed.

2. LUBRICATION EFFECT ON PIPE JACKING

In Japan, slurry pipe jacking has been firmly established as a special method for the construction of underground pipelines. Recent technological developments have produced successful methods for stabilizing unstable strata by means of employing lubricant around the pipes. A reduction in volume contraction of a lubricant after injection into the over-cutting area is needed for slurry pipe jacking. Moreover, a pressurized lubricant functions as an important support for mitigating overburden pressure.

Moreover, during the pushing processes, the lubricant is injected into the face and the over-cutting area, which is between the concrete pipe and the soil. After the lubricant fills the voids, the soil stabilizes due to the lubricant pressure during the construction work. Figure 1 shows an illustration of the roles of a lubricant.

Lubricants for slurry pipe jacking are sold in Japan, however, there has been no research done concerning the satisfying conditions needed for injection of lubricants into an over-cutting area, what defects lubricants have, and so on.

3. SURFACE-ACTIVE AGENT VISCOTOP

One of the most effective surface-active agents with flyash is called Viscotop. This surface-active agent is composed of two major components that are known as Viscotop 100A and Viscotop 100B. The chemical composition of the first part is alkyl arylsulfonate and of the second part is alkyl ammonium salt. Viscotop 100A is basically a base having a pH of between 8-10 and Viscotop 100B is an acid having a pH of between 4-8.

Adding Viscotop to flyash cement can change the properties of the final product. Ordinarily, the same quality of Viscotop 100A and 100B are used as surface-active agents. The quantity of this surface-active agent is selected as 0.5-5 percent of the weight of water in the mixture.

4. STABILITY OF FLYASH SURFACE-ACTIVE AGENT MIXTURE

4.1 Bleeding Test

Suspended particles, which are flyash and cement particles, settle in fluid at rest under the action of gravity with a velocity that is proportional to the square of the particle diameter. Coarse particles settle first, followed by finer ones, and the density of the sediment decreases with the size of the particles. Very fine particles are subject not only to the force of gravity, but also to mutually acting electrochemical forces and a Brownian motion, which appears in the suspensions of colloidal particles. Therefore, the velocity of sedimentation becomes lower than the one corresponding to the Stocks’ law [4]. The stability of the flyash and cement mixture is an important property both during and after injection. The reasons are that in the case of unstable suspension, not only are the transport and injection pipe clogged with sediment particles, but also the effect of filling the pores is decreased due to the shrinkage of the backfill mixture volume. So, in order to evaluate the stability of the flyash surface-active agent mixture, a bleeding test was carried out [5].

Flyash, surface-active agent and water are mixed and suspended in water to yield suspension samples of the desired composition and density. The flyash, surface-active agent and water are then poured into a 1,000cm³ laboratory jar and left to rest. At selected intervals of time, the sedimented volume of the denser suspension under the clear water was recorded. In this research, the height of the interface between the clear water and denser suspension indicated the sedimented volume. The height of the laboratory jar was 30cm [6]. Table 1 lists the composition of the flyash, surface-active agent and water for the samples. The components of flyash were 200, 300, 400 and 500kg per 1 m³ of each sample. The amounts of surface-active agent or Viscotop were 2, 3, 4 and 6% per 1 m³ of a solvent or water.

Figure 2 shows the relationship between the bleeding ratio one month after setting the test and the amount of Viscotop added. No bleeding was shown in the samples of C and D. Conversely, severe bleeding was observed in the samples of A and B. Since samples of A and B have few flyash particles within a unit volume, it is easy to sediment because the restraint between particles is inferior. From this point of view, it is understood that bleeding can be controlled by the amount of Viscotop added. Moreover, in the case of the 3% addition of Viscotop, there is little bleeding in each combination with regards to the amount of addition of flyash.

4.2 Slump Test

Considering the application of the mixture with flyash and Viscotop as a lubricant, testing of the viscosity or the fluidity characteristics was carried out by the slump test. The slump test is defined as the Japanese industrial standard. The dimension of the flow cone device was 51mm x 50mm. After the mixture was set into the cone on the glass board, the cone was pulled up from the original position. The mixture spread out after pulling the cone up from the original position, the length of the mixture being measured as the flow value. The flow value had good correlation with the yield stress, which is the beginning stress of the flow. Therefore, the larger the flow value, the better the fluidity.

Figure 3(a), (b) shows the relationship between the flow value and the amount of Viscotop added, where samples D and C except C-2 could not measure the flow values one month after mixture. Compared with these figures, it is clear that the fluidity of samples A and B is increased after mixture. Also, the fluidity of sample C with 3% of Viscotop being added was kept.

Taking into account these data, samples B-2 and C-2 are deemed superior for use as a lubricant because no bleeding was shown and good fluidity was kept one month after setting the test for these samples. From the cost comparison, since the amount of Viscotop in C-2 was more than that in B-2, the sample of B-2 is better because the price of Viscotop is very expensive. However, the bleeding ratio of B-2 is inferior to that of C-2. It is a suitable composition for the sample C-2 without segregation of blockades on the pump due to bleeding. Accordingly, it is concluded that samples C-2 are most superior for use as a lubricant.

5. DYNAMIC FRICTION FORCE TEST

Analysis of recorded thrust data from different pipe jacking work sites shows the predominant beneficial effects of lubrication on frictional forces. By injecting lubricant around the pipes, the jacking force is reduced by two concurrent actions: first the reduction in friction angle for the soil-pipe interface and second, the prevention of ground convergence around the pipes [7]. There are a wide range of lubrication compositions, which are empirically adjusted at worksites. The selection of a particular lubricant is mainly based on the soil condition and technical methods to be applied. This selection is easily done for small diameter pipe jacking over short distances. For large diameter and long distance jacking, however, a reliable method is needed [8]. To maintain the efficiency of lubrication over long jacking distances, the lubricant must maintain its water content [9]. This means that the chemical structure of the lubricant must be stable against dynamic friction at the pipe surface during jacking.

In order to study the effect of lubrication stability on the dynamic friction force around the pipes, a test of shear resistance at the pipe-lubrication interface was performed. For this test, flyash-surface-active agent C-2, Lubricant α and β for reference were selected. Lubricant α is basically a mixture of two parts: agent A (Na₂SiO₃) and agent B (KHCO₃). Lubricant β consists of liquid high polymer.

During the test, the frictional resistance at the interface was recorded over time. Figure 4 (a) shows the test device. The test device is composed of two container parts, namely, lower and upper parts, a connecting wire and a recorder. The lower part is filled with soil, whose components are listed in Table 2, on the bottom and lubricant material on the top. The soil layer allows the effect of the soil-lubricant interface to be effective. The upper part is filled with concrete and placed on the lower part as illustrated in Figure 4 (b). By connecting the wire to the upper part, the test is started with a rate of 1 rotation per 6 minutes.

Table 3 lists the composition of each sample. When Lubricant α and β is injected into the over-cutting area, these lubricants mix with the mud slurry. With this in mind, we prepared Lubricant α and β in these trading combination and added mud slurry of 10, 20 and 30 weight % behind the weight of each lubricant. For a reference, we also prepared only soil and mud slurry instead of a lubricant, respectively. Table 4 lists the composition of mud slurry.

Figure 5 shows the relationship between the elapsed time after starting the test and the coefficient of friction for all samples. It is understood that the coefficient of friction is about 0.6 when the lubricant into the over-cutting area is disappeared and the concrete pipe comes into contact with the surrounding soil directly. The coefficient of friction of only Lubricant α and Lubricant α added 20% mud slurry were lowest and maintained during the test. Since Lubricant α has a gelatinous solid texture, it has strength against the weight of the upper part of container. So, no failure or seepage into the concrete occurred. Conversely, the coefficient of friction of only mud slurry and Lubricant β were increased as time elapsed after starting the test. This is the reason why the mud slurry and Lubricant β easily permeated into the soil. Moreover, the added percentage of mud slurry in Lubricant β was increased, and the coefficient of force was also increased. From this point, it is understood that Lubricant β is inferior as a lubricant for the over-cutting area.

The coefficient of friction of the flyash surface-active agent mixture had converged to 0.4 except at 7 minutes after the test commenced. Also, it was found that few volume changes were observable in the flyash surface-active agent mixture due to the lack of dehydration as a result of the weight of the upper part of the container during the test. Moreover, the coefficient of friction is 0.45 or less for this kind of soil, as proposed by the Japan Society for Trenchless Technology [10]. To summarize, the coefficient of friction of the flyash surface-active agent mixture is satisfied.

6. PRESSURIZED SEEPAGE TEST

It is clear that an increase in a lubricant’s strength is a consequence of the volume change which occurs with the dehydration of a lubricant’s ingredients [11]. Accordingly, the extent of strength deterioration after injection of the lubricant into the over-cutting area can be evaluated by measuring the quantity of water exuded from the lubricant. It is not desirable that the strength of a lubricant deteriorate in the over-cutting area due to a reduction in the lubricant volume, since water percolates from the lubricant under overburden pressure and under low ground water conditions. Thus, the pressurized seepage test was carried out to evaluate the volume of percolated water obtained from the lubricant under pressure.

Figure 6 shows the pressurized seepage test procedures. An acrylic column 75mm in diameter and 300mm in length was prepared. After this column was filed with distilled water, glass beads with a diameter of 1mm were thrown into the column to a depth of about 150mm. Then, the surplus distilled water displaced by the glass beads was discharged. 250ml each of the flyash surface-active agent mixture, Lubricant α and β, were prepared, and poured into the glass bead layer immediately after mixing, and the pressurized seepage test was carried out 10 minutes after each lubricant was prepared. 0.5kPa of compressed air was applied to the lubricant, and the weight of percolated water from the glass bead layer and the dehydrated water from a lubricant was measured.

Figure 7 shows the relationship between the amount of percolated water and the elapsed time. It was found that the amount of percolated water from the flyash surface-active agent mixture was minimal. Conversely, the amount of percolated water from Lubricant α was 1.5 times that of the flyash surface-active agent mixture. However, each lubricant increased regarding the amount of accumulated percolated water, with the nonlinear curve. The structure of a lubricant in a gelatinous state is fundamentally weak, and it becomes so as a result of dehydration where percolating water is discharged from the clearance part of the gel by applying pressure to a lubricant. Moreover, the extent of the discharge of percolated water is considered to change with differences in the degree of breakage of a structure. Since the percolated water, by which pressurization dehydration was carried out from the lubricant, permeates the surrounding soil, it easily causes the collapse of an over-cutting area with the reduction of the volume of a lubricant, and increases the thrust and subsidence in Lubricant β. Therefore, it is thus found that Lubricant β is inferior as a lubricant for use on the over-cutting area.

7. LONG-TERM IMMERSION TEST IN DISTILLED WATER FOR LUBRICANT α

If soil is fully saturated, the function of a lubricant may be lost due to elution of components of a lubricant [12]. The main component of Lubricant α is a silicate, and it is well-known that strength deterioration may occur by an electrolyte being exuded [13]. Therefore, a long-term immersion test in distilled water was carried out in order to evaluate the long-term stability of the lubricant in underground water.

After Agents A and B of Lubricant α at a 1/2,000 scale per combination of 400 liters for each agent were mixed, it was poured into a mould 5cm in diameter and 10.5cm in height. It was kept in the mould 10 minutes after mixing and was then removed. It was set in the center position of a beaker into which 900 ml of distilled water was poured. The distilled water in the beaker was taken out by siphon after 24 hours, and the distilled water was poured in after the sample was photographed. This procedure was repeated every day for two weeks.

Figure 8 shows the situation of immersion progress of the Lubricant α with the least stability, Lubricant E. It was found that the strength deterioration of Lubricant α was recognized four days after immersion. Half the volume of it broke eleven days after immersion. Furthermore, it broke into pieces 14 days after immersion.

As for the lubricant α of a silicate, it is generally considered that the strength of it does not change and the stability of its water resistivity is high. However, the strength deterioration is clearly recognized by the elusion of components of the lubricant such as sodium and potassium. From these perspectives, Lubricant α may deteriorate easily in a saturated soil. This means that a problem exists regarding long-term durability. As a result, volume contraction of Lubricant α after injection into the over-cutting area occurs easily. Moreover, the important role as the support for the overburden pressure may be compromised.

8. ION ELUTION FROM FLYASH SURFACE-ACTIVE AGENT MIXTURE

The environmental restriction value of hexavalent chromium is set to 0.05 ppm or less [14]. From this point of view, quantities of elution hexavalent chromium ions from the flyash surface-active agent mixture have to be clarified. The tank leaching test was carried out for an investigation of this property. Table 5 lists the components of the sample.

Figure 9 shows the relationship between elapsed time and the quantity of elution of the hexavalent chromium. Although the quantity of chromium in Figure 9 is treated with a fixed quantity of the total amount of chromium, since trivalent chromium turns into insoluble hydroxylation chromium in alkali slurry conditions which use hydraulic powder, this brings the same result as having treated it with a fixed quantity of only the hexavalent chromium. It can be said that the flyash surface-active agent mixture did not elute hexavalent chromium. Taking this into account, the flyash surface-active agent mixture is excellent for controlling the elution of the components.

9. CONCLUSIONS

In this paper, in order to examine the characteristics and behavior of the flyash surface-active agent mixture material developed as a lubricating material, a fundamental investigation was performed. The following conclusions can be drawn from this study:

1. In the case of the 3% addition of Viscotop, there is little bleeding in each combination with regards to the amount of addition of flyash.
2. The coefficient of friction of the flyash surface-active agent mixture was converged to 0.4.
3. The amount of percolated water from the flyash surface-active agent mixture was minimal based on the pressurized seepage test.
4. The flyash-surface active agent mixture is excellent for controlling the elution of components such as hexavalent chromium.

Taking these data into consideration, the flyash surface-active agent mixture is determined as being excellent for use as a lubricant. 

REFERENCES

[1] Shimada H. and Matsui K., 1998, A new method for the construction of small diameter tunnels using pipe jacking, Proceedings of Regional Symposium on Sedimentary Rock Engineering, 234-239.
[2] Pellet-Beaucour A. L. and Kastner R., 2002, Experimental and analytical study of friction forces during microtunneling operations, Tunneling and Underground Space Technology, Vol.17, 83-97.
[3] Yoshida Y., Shimada H., Gottfried J., Iino M., Matsui K., Ichinose M., Fujita S., Mineshita M., Mizunuma T., Kawai T. and Sakai E., 2006, Application of injection material for rehabilitation of pipeline using flyash-surfactant mixtures , Proceedings of 10th Conference on Environment and Mineral Processing , Ostrava, Czech, Vol.1, 93-100.
[4] Nonveiller E., 1989. Grouting Theory and Practice.
[5] Shimada H., Sasaoka T., Kubota S., Matsui K. and Yoshida Y., 2003, The application of flyash cement in mining backfill material, Proceedings of International Symposium on Mine Planning and Equipment Selection, Kalgoorlie, Australia, 199-204.
[6] Shimada H., Sadisun A. I., Ichinose M., and Yoshida Y. and Matsui K., 2003, The application of flyash cement as lining material for slope stability of slake-prone rocks, Proceedings of the 4th International Conference on Computer Applications in the Minerals Industries, Calgary, Canada, CD-ROM.
[7] Barla M., Borghi X., Mair R.J. and Soga K., 2003, Numerical modelling of pipe-soil stresses during pipe jacking in clay’, Nota accettata al XIIIth ECSMGE, 25-28.
[8] Khazaei S., Shimada H., Kagawa K., Kawai T., Yotsumoto J. and Matsui K., 2004, Method of changing damaged pipes in curve area of pipe jacking -Application in a slurry pipe jacking project in Tokyo-,. Proceedings of International NO-DIG 2004 Conference, Hamburg, GERMANY, CD-ROM.
[9] Khazaei S., Shimada H., Kawai T., Yotsumoto J., Sato I. and Matsui K., 2003, Laboratory consideration about frictional force on pipe surface when slurry machine is used, Proceeding of International Symposium on Fusion Technology of Geosystem Engineering, Rock Engineering and Geophysical Exploration, 214-220.
[10] Shimada H., Kawai T., Sasaoka T., Matsui K., Ichinose M., Fujita S. and Yoshida Y., 2006, Behavior of Lubricant Injection into Over-cutting Area for Slurry Pipe Jacking, Proceedings of International NO-DIG 2006 Conference, Brisbane, Australia, CD-ROM.
[11] Ishibashi N., 1998, Jacking Method –Technology and Cost Reduction-, Sewage Works in Japan.
[12] Khazaei S., Shimada H., Kawai T., Yotsumoto J. and Matsui K., 2006, Monitoring of Over Cutting Area and Lubrication Distribution in a Large Slurry Pipe Jacking Operation, Journal of Geotechnical and Geological Engineering, Vol.24, No.3, 735-755.
[13] Khazaei S., Wu W., Shimada H. and Matsui K., 2006, Effect of Lubrication Strength on Efficiency of Slurry Pipe Jacking, Proceedings of GeoShanghai Conference, Shanghai, China, 170-177.
[14] Akima T., 2003, The Chromium LEACH Cause from Cement Improvement Soil in Specific.

¹ Kyushu University, Fukuoka, Japan
² Kyowa Exeo Co. Ltd., Tokyo, Japan
³ Sankyo Material Co. Ltd., Fukuoka, Japan