Comparative study on foundation treatment methods of immersed tunnels in China

Abstract

Based on engineering practices of four typical traffic immersed tunnels built in China, this paper details the features of the four dominant foundation treatment methods for immersed tunnel construction: pile foundation, sand flow foundation, grouting foundation, and gravel bedding foundation. Subsoil stress time-history of different method are specified first, plus a summary of settlement assessment method for foundation quality control. Further, a comprehensive comparison of settlement and cost of these four foundation treatment methods is conducted to highlights the specific merits, disadvantages and conditions encountered in each foundation treatment method, based on real projects information. The findings of this article could henceforth be applied to foundation treatment work in immersed tube tunnel construction.

Keywords

foundation treatment method / immersed tunnel / subsoil stress / settlement

Introduction

Immersed tunnels refer to tunnels built with special immersion method. They are usually under waterways and considered better than other crossings like bridges or bored tunnels. Tunnels are superior to bridges mainly in that they do not disturb ship navigation, especially for busy water channels. Technically, bored tunnels must have a minimum safety-buried-depth and thus are usually constructed at a certain depth below the river bed, which inevitably increase the whole tunnel length under limited longitudinal slop. Immersed tunnel structurally consists of factory-made elements (usually about 100 m long) which are connected underwater with special rubber gaskets. Notably, immersed tunnel can be directly placed at a pre-excavated trench on the shallow river bed, and no need for minimum burial depth (unlike bored tunnel), so when at the same construction site, the total length of an immersed tunnel generally can be shorter than that of a bored tunnel [¹]. Immersed tunnels are mostly constructed under canals and waterways, especially where no ship navigation interference is allowed. Until the end of 2018, there are more than 200 immersed tunnels in the world, and most of them are in the North America, Europe, and Asia. For example, in the Netherlands alone, more than 30 immersed tunnels are in-service. What’s more, more immersed tunnel projects, including the 6.7 km Shenzhen-Zhongshan Bridge Tunnel, the 18 km Fehmarnbelt fixed road and rail link, et al. are being constructed or designed for fixed links beneath waterways [²].

Although immersed tunnel has a quite long history of over 100 years, the relevant techniques were first introduced into mainland China in the late 1980s, and the recent 30-year has witnessed a series of immersed tunnels being finished, some even marked the milestone in tunnel history. The first immersed tube tunnel in China mainland was completed in 1993. Figure 1 presents the overviews of the cross section, usage, dimension and maximum depth of 11 immersed tunnels finished in China.

Generally, immersed tunnel construction will first excavate a trench on riverbed and a specially prepared foundation is made before tunnel body is finally placed on. Foundation treatment always plays an important role in the construction of immersed tunnels because dredging always tends to leave an uneven bottom with unacceptable tolerance for the accuracy of tunnel alignment. With so many tunnels being built, it is important to have a comprehensive understanding of foundation treatment methods based on comparative study of past engineering practices. Although some researchers have studied the features of each foundation treatment method, relevant summary remains very limited. For example, development of foundation treatment methods in Holland since the Mass tunnel is summarized and explained in detail [³]. As a main factor of tunnel settlement, the method of tunnel foundation construction is discussed based on 15 tunnel projects in history, but no detailed of project geological conditions are provided, nor is the comparison study between different method specified [⁴,⁵]. With a 30-year development, immersed tunnel projects in China have provided very good cases for such an important comparison study. Table 1 summarizes the foundation treatment methods of the built 11 immersed tunnels in China. Four typical foundation treatment methods have been applied in these tunnels (pile, grouting, sand flow, and gravel bedding), and these four methods represent the typical immersed tunnel foundation treatment methods currently widely used around the world. In China, all these foundation treatment methods have been explored and applied in real projects, which provide valuable experience and reference for tunnel engineering.

This paper first summarizes and compares the selected four foundation treatment methods, including analysis and explanation of foundation settlement from soil stress time-history perspective, and the settlement calculation method. Secondly, each method is further explained with a specific tunnel project in China, from aspects of geotechnical conditions, foundation design, construction techniques, settlement performance, and other issues. With the aim of providing reference value for future immersed tunnels, experiences and lessons of these methods concluded from engineering practices in China are finally discussed in this paper.

Development of foundation treatment method

An overview of the applications of foundation treatment methods

By reviewing the foundation treatment methods for immersed tube tunnels in the world, there has been a tendency historically to associate gravel bedding foundation method with steel tunnels in America, sand flow foundation method or sand jet foundation method in Europe and grouting foundation method in Japan. However, this association is not fully valid. In each particular case, the foundation treatment method should be chosen with fully consideration of the local geology, environment and other specific conditions. A summary of the selected four typical foundation treatment methods together with a method comparison table (Table 2) are provided below.

Four types of foundation treatment methods

Sand flow foundation

Sand flow method was developed from sand jetting method and then first used in Vlake Tunnel of Holland in 1975 [⁶]. Compared with sand jetting method, the slowly-moving gantry which is the main disadvantage in the sand jetting system is eliminated in this new method so that there is no hindrance to navigation at all. To shorten the supply pipes in long tunnels, the sand flow method was then improved in the Hemspoor Railway tunnel (1980) which feeds the sand-water mix through the pre-embedded pipe in the floor slab from the outside floating equipment. However, the sand flow method was applied for the first time at the Zhujiang Immersed Tunnel (1995) in China.

Sand flow method is used to create a sand foundation by pumping the sand-water mix into the gap between the underside of tubes and the bottom of trench through a pipe system with opening on the bottom of the tubes. As the velocity of the mix decreases after pumped out of the injection point, sand settled forming a circular pancake around the injection point. These injections points are usually designed at an approximately distance of 10–15 m from each other. By injecting the mixture point by point, a pattern of overlapped sand pancake is gradually formed to fully fill the gap and make a firmly foundation beneath the tubes.

Pile foundation

The first application of pile foundation treatment method can be traced back to 1912 with the completion of La Salle St. Tunnel [⁷]. Pile foundation method is mainly used in special cases when a direct foundation on the subsoil will be impossible in practice with regards to the settlement and bearing capacity. For example, Rotterdam Metro Tunnel which is completed in 1966 applied this method due to large dynamic train load and varied stiffness of the subsoil. The Changhong Tunnel built in 2002 first applied pile foundation in China with the aim of reducing settlements in serious siltation condition.

When applying a pile foundation, the most important work is to fill the gap between the pile top and the tunnel bottom, since it is impossible to keep all the pile heads at the same level. To eliminate the gap between the pile head and the bottom of the tubes, several adjustment methods have been developed in the tunneling history. These methods could be divided into two types: ‘adjustable pile head method’ and ‘pile capping beam method’. The first type is usually applied with an adjustable head consisting of a separate concrete part connected to the rest of the pile by a nylon sleeve. After the tube is immersed, the adjustable pile head will be uplifted to the bottom of the tubes by grouting into the nylon sleeve. The adjustable pile head method is first applied in the foundation construction of the Rotterdam Metro Tunnel. The second method is usually applied with concrete or steel pile capping beams constructed underwater to hold a group of piles together. The gap between the capping beams and the tubes will be usually filled with cushion material and cement grout. It’s worth mentioning that a grouting bag method was developed and applied in the Changhong Tunnel which would be further explained in Section 3.2.2.

Grouting foundation

Grouting foundation method is first applied in the Tokyo Port Dainikoro Tunnel in 1976. This method was then widely used in Japan characterized by solving the liquefaction problem of foundation constructions in seismic area. This method was applied in China in three tunnels: Yongjiang Tunnel (1995), Haihe Tunnel (2011), and Shenjiamen Harbor Tunnel (2014).

Grouting foundation method is developed from filled bag method with the advantages of eliminating the need of costly bags and avoiding the complicated fixing technique requiring diving and water-borne manipulation. After the excavation of the trench, a layer of crushed stone should be laid on the bottom of the foundation trench. After the element is immersed on the temporary support, a riprap of approximately one meter in height should be placed along the element’s side wall and rear bulkhead to seal off the bottom of the element. From embedded grouting holes within the element’s bottom slab, grout is then injected into the voids between element and its gravel bedding to form a good support foundation.

Gravel bedding foundation

The traditional gravel bedding foundation is first applied in the Detroit Windsor Tunnel in 1930. Since then, this traditional gravel bedding foundation method is typically used in steel immersed tunnels with octagon or circular cross-sections. However, the traditional gravel bedding foundation method is not suitable for the concrete immersed tunnel with wide rectangular cross-section because of the great difficulties in construction. To deal with the complex off-shore condition and achieve higher accuracy of leveling tolerances, the Øresund Tunnel Contractors (ØTC) successfully applied an improved gravel bedding foundation method in the Øresund Tunnel built in 2000 [⁸,⁹]. After the application of Øresund Tunnel, this improved gravel bedding foundation method has been further applied in more off-shore immersed tunnels, such as the Busan-Geoje Tunnel and the HZMB Tunnel [¹⁰].

The basic concept of the improved gravel bedding foundation method is to construct a rubble stone of an intermittent pattern of identical berms alternated by grooves instead of a closed plane [⁸]. Gravel was transported down to the trench through a fallpipe which is attached to a special designed vessel. The lower end of the fallpipe is equipped with a screeding plate so that placing and screeding of the gravel is executed at the same time. The lower end of the fallpipe is fixed at a constant level to placing the gravel to required level, while moving horizontally. With different kinds of measuring systems and compensation measures, the accuracy of the level tolerances can be achieved within centimeters.

Subsoil stress time-history and settlement assessment of different foundation treatment methods

As settlement control is an important factor in deciding the best foundation treatment, it is meaningful to analyze the subsoil stress time-history during the immersed tunnel construction and service periods. According to the published settlement data of over 30 immersed tunnels, it is reasonable to divide the total settlement into two parts: settlement in construction period and settlement in service period. The former is mainly related to the soil property and construction procedures, while the latter is mainly related to consolidation of subsoil and additional loading effects from riverbed sedimentation. Figure 2 shows the subsoil stress time-history through the whole tunnel construction stage. First stress decrease of \mathrm{\Delta }{p}_{\mathrm{1}} (shown by oa) is due to the trench excavation before foundation treatment, this leads to a significant unloading on the subsoil. It should be noted that this trenching process usually causes relatively big bounce-back on the trench bottom, especially for immersed tunnel with large bury depth, for example, the trenching depth of Hong Kong-Zhuhai-Macao Immersed Tunnel is designed to be as high as 40 m. After trenching, usually the excavated trench bottom will be paved with a sand or gravel layer for primary levelling, then re-loading effects will come out along the preceding construction schedules.

Different foundation treatment methods will display different re-loading curve. For gravel bedding, see the dotted line abcde in Fig. 2, the gravel bed is formed once so as to trigger a gradual loading (line ab) in subsoil, then tunnel element is directly immersed underwater and put on the prepared gravel bed layer, this further increase the subsoil stress, as shown by line bc. After that locking backfill on both sides and protection backfill on the top continues the loading effects, and riverbed sedimentation in service period generally further causes subsoil loading, as shown by the line cde. For grouting and sand flow foundation treatment, the re-loading process present some differences shown in the solid line afmnps, and this is due to the quite different construction procedures. After trench excavation, the bed is first paved with a thin fine-gravel cushion layer, which cause a minor loading (shown by line af), then the tunnel element is placed on a temporary support (piles) above the cushion layer, and then the gap between them are filled by sand (in sand flow method) or by cementitious grouting (in grouting method), as shown by fm. after that element is placed on the sand or grouted layer (mn), and finally, much similar to the gravel bedding, backfilling and riverbed sedimentation further increase the loading effects. It should be pointed out that riverbed sedimentation is the main cause of excessive settlement in many in-service immersed tunnels.

The foundation settlement can generally be calculated based on the e-logp plot method (Fig. 3). The subsoil should be carefully investigated and sampled, so as to get the detailed consolidation property. The subsoil is modeled as over-consolidated soil, and unloading-reloading should be reflected for higher accuracy. The ground is usually subdivided into several distinguished layers with largely different parameters, and final settlement is calculated by layer-wise summation method as below,

(1)s={\displaystyle \sum _{i=\mathrm{1}}^n\frac{\mathrm{\Delta }{e}_i{H}_i}{\mathrm{1}+e_{\mathrm{0}i}}},

where H_iis the thickness of soil layer i, e_{\mathrm{0}i}is the initial void ration of soil layer i, \mathrm{\Delta }{e}_{\mathrm{0}i}is the void ration change of soil layer i.

When \mathrm{\Delta }{p}_{\mathrm{2}i}+\mathrm{\Delta }{p}_{\mathrm{3}i}+\mathrm{\Delta }{p}_{\mathrm{4}i}\ge p_{\mathrm{c}i}-p_{\mathrm{o}i}, then

(2)\mathrm{\Delta }{e}_i=C_{\mathrm{s}i}\log {\displaystyle \frac{p_{\mathrm{c}i}}{p_{\mathrm{0}i}}}+C_{\mathrm{c}i}\log {\displaystyle \frac{(p_{\mathrm{0}i}+\mathrm{\Delta }{p}_{\mathrm{2}i}+\mathrm{\Delta }{p}_{\mathrm{3}i}+\mathrm{\Delta }{p}_{\mathrm{4}i})}{p_{\mathrm{c}i}}}.

And when , then

(3)\mathrm{\Delta }{e}_i=C_{\mathrm{s}i}\log {\displaystyle \frac{(p_{\mathrm{0}i}+\mathrm{\Delta }{p}_{\mathrm{2}i}+\mathrm{\Delta }{p}_{\mathrm{3}i}+\mathrm{\Delta }{p}_{\mathrm{4}i})}{p_{\mathrm{c}i}}},

whereC_{\mathrm{c}i},C_{\mathrm{s}i} are the compression index and expansion index of soil layer i, respectively; p_{\mathrm{o}i} is the average vertical effective stress (the overburden stress) of soil layer i;p_{\mathrm{c}i} is the average pre-consolidation stress of soil layer i; \mathrm{\Delta }{p}_{\mathrm{2}i}/\mathrm{\Delta }{p}_{\mathrm{3}i}/\mathrm{\Delta }{p}_{\mathrm{4}i} are the increase of effective stress in soil layer i during foundation treatment, tunnel element placement and backfilling.

It should be noted a careful geological investigation work is critical to obtain necessary parameters for foundation quality evaluation and for settlement assessment. Sometimes uncertainties in soil parameters should be considered for better analysis.

Case histories of foundation treatment methods in China

Shanghai Outer Ring Tunnel Project

Introduction

Shanghai Outer Ring Tunnel is built by seven concrete elements with a wide rectangular cross-section of 43 m× 9.55 m each. This 2882 m long highway tunnel consists of three tubes for eight lanes and two service galleries for rescue and cable shown in Fig. 4. The immersed section part of this tunnel is 736 m. Figure 5 shows the geotechnical and geomorphic profile along the tunnel alignment. The terrain of the tunnel site slopes abruptly in the west and gently in the east. The subsoil along the tunnel alignment is mainly composed by mucky soil, silt clay, and silt clay with sand. There is a deep trough beneath the river at a depth of 23.6 m making the riverbed an asymmetry V type. The seismic intensity at this tunnel site is 7.0 in accordance to Chinese design criteria.

Construction techniques of sand flow foundation

Sand flow foundation method is applied in Shanghai Outer Ring Tunnel. Geological condition and sand-water mixture supply system of Shanghai Outer Ring Tunnel are two distinguished aspects compared with the Zhujiang Tunnel which is the first immersed tunnel applying the sand flow foundation method in China. Shanghai Outer Ring is mainly placed on a soft foundation, while Zhujiang Tunnel is mainly placed on an intermediary weather rock foundation. Furthermore, the sand-water mixture is supplied not from inside but from outside sand filling ship through pipelines pre-embedded in the slab of elements (Figs. 6(a) and 6(b)). This supply system is of advantage to shorten the length of supply pipelines and eliminate the one-way ball valves.

The injection holes are arranged at distances of 10 m in the slab, 4 in each row horizontally. Cement clinker is added to the sand at a ratio of 0.05:1 to prevent the liquidation under the seismic intensity of 7 degree. The radius of the sand deposit is designed as 7.5 m. Other relevant construction parameters are also shown in Table 3. For the convenience of siltation removal, the injection should start longitudinally from the first row to the penultimate row of each element leaving the last row non-injected and the last element should be injected in the opposite sequence to elements ahead. Within each row, the injection should start from the central holes and proceed toward to both sides. After the process of sand-water mixture injection is finished in each element, the injection holes are grouted by pressure again to fill up craters and other small Interstices.

The quality of sand deposits was verified through sand quantity control, pressure analysis, diving investigation and element elevation monitoring. To calculate the sand quantity for injection, the depth of the trench must be surveyed before immersion. The pump discharge pressure should not exceed 0.1 MPa. To prevent the element uplift, the injection would be stopped as soon as the uplift value of the element exceeded 5 mm or the single jack pressure of temporary foundation reduced to 1400 kN.

Settlement Analysis

The accumulated settlement curve of the tunnel since 2003 is shown in Fig. 7. From these measuring results, the largest settlement occurred at the joint between E5 and E6 reaching to 230 mm. The annual accumulated settlement at the joint between E5 and E6 in 2004 reached to 120 mm accounted for 50% of the total settlement. The central part of E7 experienced a relatively large settlement of 40 mm in 2007. The largest differential settlement between both ends of the element occurred at E6 reaching to 172 mm. By grouting into foundation under the E5/E6 joint and E7 in 2005 and 2007, respectively, the settlements at these critical locations have been controlled.

Ningbo Changhong Tunnel

Introduction

Ningbo Changhong Tunnel built in 2002 is composed of four concrete elements with a rectangular cross-section of 22.8 m× 8.45 m each. The total length of this tunnel is 3541.15 m consisting of the 897.24 m long north section, the 395 m long immersed section and the 2248.91 m long south section. Figure 8 shows the geotechnical and geomorphic profile along the tunnel alignment. The minimum of overburden and largest gradient of this tunnel is 1.5 m and 4%, respectively. The subsoil is mainly composed of mucky clay, silt clay, and silt clay with sand belonging to the Quaternary system. The seismic intensity at this tunnel site is 7.0 in accordance to Chinese design criteria.

Construction techniques of pile foundation

Considering the serious siltation of approximately three meters per month in thickness on such soft subsoil, pile foundation method was applied in this tunnel. The tunnel was supported by 204 driven 600 mm × 600 mm piles and 16 bored piles at distances ranging from 5 to 8.65 m. These piles were arranged four in each row (Fig. 9(a)). The clearance between the bottom of tube and the top of the pile is filled up by using grouting bags. The fully-filled grouting bag is 400 mm in thickness and 1.5 m in diameter. The grouting bag together with the grouting pipes is pre-imbedded inside the tube bottom at the position of its corresponding pile. For the convenience of chopping piles underwater, the upper 3 m section of the pile is made of steel at a diameter of 750 mm, which is connected to the below pre-stressed RC pile through flanges. A steel plate which is 800 mm in diameter is then welded on the top of pile (Fig. 9(b)).

After immersion of the tube, a clearance of approximately 200 mm between the bottom of the tube and the top of the pile would be left. After fully injected, these grouting bags filled the gap and connected the element and piles so that the upper load could transmit to the pile foundation. The gap between the bottom of the tube and trench should be finally grouted to form an overall foundation together with these square piles.

To control the quality of the grouting bags, the grouting bags should always be completely injected. The grouting pressure should not exceed 0.05 MPa than the water pressure at this position with a suggested range of 0.02–0.04 MPa. Besides, upon the completion of the grouting bag injection, divers should be appointed to verify the bags are solidly connected to the top of the piles.

Compared with the adjustable pile head method and pile capping beam method, the grouting bag method is featured by simple fabrication, steel or concrete material saving and diving work reducing. However, the accuracy requirement for pile driving and tube immersion is higher than the two previous methods.

Settlement analysis

Pile foundation method effectively reduces the accumulated settlement of the Changhong Tunnel. The accumulated settlement of this tunnel is only 20 mm after two years operation. Table 4 shows that the largest accumulated settlement is only 6.7 mm within period from September 2006 to December 2009 (Table 4).

Ningbo Yongjiang Tunnel

Introduction

The Ningbo Yongjiang Tunnel is built in 1995 by five elements with a rectangular cross-section of 11.9 m × 7.65 m. The immersed tube section of this 1020 m long tunnel is 420 m in length. Figure 10 shows the geotechnical and geomorphic profile along the tunnel alignment. The subsoil at this tunnel site is mainly composed of loose sediments belonging to the Quaternary system which is very similar to the Changhong Tunnel. The tunnel is mainly located in the upper saturated mucky soil layer which is about 20 m in thickness. The measured standard penetration value of this mucky soil layer is only 1–2.

Construction techniques of grouting foundation

Grouting foundation treatment method is applied in the Yongjiang Tunnel. The trench was excavated by 8 m³ grab dredger. During the trench dredging, two grade slopes were set up to ensure their stabilities and the gradients from the bottom to top are 1:3 and 1:4. Before the immersion of the element, a 60 cm-thick gravel layer was placed on the surface of the trench. After immersion, the element was supported on temporary foundation and 1.5 m lock fillings were installed at both sides of the tunnel. The grouting was then injected to fill up the 40 cm gap between the element and its gravel bedding (Fig. 11).

Based on the large-scale simulation tests, the distance between injecting holes was designed as 5.5 m and the grouting pressure was designed as 0.05 MPa higher than the local water pressure. Based on the indoor and in-site material tests, three types of additives were added to the mix proportion (water: cement: bentonite: fly ash: sand= 2:1:0.2:0.8:4.5) to improve the performance of grouting material on stability, uniformity, and pump ability. The quality of the grouting results was checked out by grouting amount statistics, observation holes, and non-destructive investigation method.

Settlement analysis

A total amount of 1807 m³ was finally grouted in this project. The accumulated settlement of this tunnel and the measuring points are shown in Fig. 12. The settlements at J1 and J6 are relatively small because the two ends of the tunnel are placed on caisson and pile foundation, respectively. Except J1 and J6, the measured data of J2–J5 shows that the tunnel settlement increased quickly especially in early operation stage. Regular dredging has been started to remove the deposition on the top of the tunnel since 1998. After 5 years’ operation, it showed that the tunnel settlement finally became steady since 2001. The largest settlement occurred at J5 reached to 88.38 mm.

Hong Kong-Zhuhai-Macao-Bridge (HZMB) Immersed Tunnel

Introduction

The Hong Kong-Zhuhai-Macao Bridge (HZMB) Immersed Tunnel is composed of 33 rectangular concrete elements, with a standard length of 180 m each. The cross-section of this tunnel is 37.95 m × 11.40 m comprising two traffic tubes and one central service gallery for ventilation and rescue. This 5990 m long tunnel consists of one 5664 m long immersed section and two 326 m long cut and over sections. Figure 13 shows the geotechnical and geomorphic profile along the tunnel alignment. The bedrock is mainly composed by mixed granite and schist belonging to the Sinian system, and covered by a Quaternary soil layer of 60–100 m thickness [¹²]. The tunnel is buried mainly on mucky soil or silt clay in both ends and on sand or silt clay in the middle. The maximum water depth of tunnel element immersion is 44.5 m.

Construction techniques of gravel bedding foundation

Considering the poor and variable geotechnical conditions along the tunnel alignment together with over 20 m thick deposited backfill on the top of the tunnel, several foundation treatment methods are applied to reduce the ultimate settlements and differential settlements of the tunnel (Table 5). Pre-stressed high-strength concrete (PHC) pile and high-pressure jet grouting (HJG) pile are applied in both two island sections, while SCP pile composite foundation is applied in both two slope sections. A 1.5 m thick, 31.95 m wide gravel bedding made of 20–75 mm sized stones is installed in the middle section of the tunnel and a 2 m thick riprap cushion layer is installed below to ensure the uniformity of the gravel bedding.

The structure and geometry of the gravel bedding foundation is presented in Fig. 14(a). The gravel is placed in discrete 1.8 m wide bunds forming a ‘Z’ type pattern in longitudinal direction. To meet the accuracy requirements of the gravel bedding construction, a homemade pontoon is developed for this tunnel (Fig. 14(b)). The specified pontoon is capable of delivering an overall level accuracy of ±35 mm from the design line throughout immunized to the complex hydrologic condition at a maximum depth of approximately of 50 m. HZMB tunnel has been finished and opened to traffic in the end of 2018. Until now the tunnel settlement information of this tunnel is unknown.

Discussions

Based on above theoretical analysis and engineering practices in China, experiences and lessons are concluded and discussed together by a comprehensive comparison in Table 6. It should be noted that since settlement and other projects information of HZMB Immersed Tunnel are unknown at present, another immersed tunnel, Zhujiang immersed tunnel, is specified here. Some basic conclusions can be reached as below:

1) Pile foundation applied in the Changhong Tunnel achieves the smallest settlement of only 2 cm in very soft ground with serious trench siltation problem, which proves better in settlement control than sand flow and grouting method in similar geological conditions. As in both sand flow and grouting, muck deposition on the foundation layer is inevitable, which will cause a major deformation risk.

2) Pile foundation method has the highest cost in the short term and lowest cost in the long term. The foundation/contract ratio (in cost) of pile foundation sums up to 5% much higher than other two methods during the construction stage. However, both Yongjiang Tunnel and Outer Ring Tunnel have encountered many problems during the operation stage: differential displacement, cracks and leakage. Additional treatment measures have to be carried out to maintain the normal operation such as regular dredging and secondary grouting which increased the total cost of these tunnels in the long term.

3) Subsoil conditions affect the performance of foundation treatment method greatly. For Outer Ring Tunnel and Zhujiang Tunnel, the sand flow treatment method applied in different subsoil conditions results in totally different settlement performance. The largest settlement in the Outer Ring Tunnel is nearly 8 times higher than that of the Zhujiang Tunnel. The settlement data of J2–J5 from Yongjiang Tunnel also indicates that subsoil consisting of thicker soft layers tends to experience larger settlement.

4) Gravel bedding method has been successfully applied in the HZMB Immersed Tunnel with an levelling accuracy of ± 35mm from the design line. When dealing with these long off-shore tunnels, gravel bedding method has the advantages of preventing liquidation, sparing temporary foundation, and harmless to concrete elements. The long-term performance of the HZMB Immersed Tunnel remains to be observed.

Conclusions

This paper makes a detailed comparison of four typical foundation treatment methods encountered in China so far in the aspects of both theory and practice. Geotechnical condition, foundation design, construction technique, and settlement performance are fully presented in each project case. Engineering experiences on foundation treatment method are discussed and concluded in the end. It’s hoped that these valuable practical experiences in China could benefit the future foundation construction of immersed tunnels in the world.

References

[1] Lunniss R, Baber J. Immersed Tunnels. New York: CRC Press, 2013

[2] Pedersen S K, Brøndum S. Fehmarnbelt fixed link: The world’s longest road and rail immersed tunnel. Civil Engineering (New York, N.Y.), 2018, 171(5): 17–23

[3] Van Tongeren I H. The foundation of immersed tunnels. In: Proceedings of Delta Tunneling Symposium. Amsterdam, 1978, 48–57

[4] Grantz W C. Immersed tunnel settlements. Part 1: Nature of settlements. Tunnelling and Underground Space Technology, 2001, 16(3): 195–201

[5] Grantz W C. Immersed tunnel settlements: Part 2: Case histories. Tunnelling and Underground Space Technology, 2001, 16(3): 203–210

[6] Glerum A. Developments in immersed tunnelling in Holland. Tunnelling and Underground Space Technology, 1995, 10(4): 455–462

[7] Rasmussen N S, Grantz W C. Catalog of Immersed Tunnels International Tunnelling Association Immersed and Floating Tunnels Working Group: State of the Art Report. 1997

[8] Smink M. Scrading—A new approach to the foundation of concrete tunnel elements. In: Proceedings of the ITA World Tunneling Congress 2003. Amsterdam: A. A. Belkema Publishers, 2003, 287–289

[9] Marshall C. The Øresund tunnel—Making a success of design and build. Tunnelling and Underground Space Technology, 1999, 14(3–4): 355–365

[10] Janssen W, de Haas P, Yoon Y H. Busan-Geoje Link: Immersed tunnel opening new horizons. Tunnelling and Underground Space Technology, 2006, 21(3): 332–340

[11] Xie X, Wang P, Li Y, Niu J, Qin H. Monitoring data and finite element analysis of long term settlement of Yongjiang immersed tunnel. Rock and Soil Mechanics, 2014, 35(8): 2314–2324 (in Chinese)

[12] Hu Z, Xie Y, Wang J. Challenges and strategies involved in designing and constructing a 6 km immersed tunnel: A case study of the Hong Kong-Zhuhai-Macao Bridge. Tunnelling and Underground Space Technology, 2015, 50: 171–177

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature